Result node message

* critical alert ui

* updating ui, backend logic

* revert

* type check fix npm

* checking out cli files from main

* moving alert icon

* adding sqlx mock data

* more sqlx changes

* feat(frontend): nodes from flow can be connected directly in expr input through a plug icon (#4652)

* Add flow prop picker

# Conflicts:
#	frontend/src/lib/components/propertyPicker/PropPicker.svelte

* fix unwanted copy

* cleaning

* Fix unset context

* move button and always display input

* fix unwanted proppicker display

* update

* update

* clean all

* clean all

---------

Co-authored-by: Ruben Fiszel <ruben@rubenfiszel.com>

* replace hide/show with toggle

* adding mutable setting and navigation to settings to configure channels

* merge fix

* ee non ee changees

* auto-acknowledge when muted

* pr comments

* fix bad log

* user inner modal component

* update unaknowledge alerts after acknowledging from modal

* aknowledge -> acknowledge

* format

* adding backend support for workspace critical alerts

* immediately check for alerts

* immediately check for alerts

* adding openapi changes

* simplify loading of superadmin/ee

* update modal logic

* frontend logic update

* sqlx prepare

* adding mute functionality for workspace critical alerts + show alerts button for instance and workspace settings

* npm / rust warnings & errors

* reverting non-mac cargo toml

* reverting non-mac cargo toml

* mute toggles in critical alert ui

* ui polish

* adding acknowledged_workspace column

* sqlx prepare

* make sure we wait for all stores to be loaded

* adding workspace mute logic to report_critical alert

* auto ack recovered alerts

* toggle workspace as superadmin and critical alerts in logs menu

* adding critical alert button if width < 786px

* refresh on change

* Move notification to logs (#4698)

* Move notification to logs

* removing unused button

---------

Co-authored-by: Alexander Petric <petric.al@gmail.com>

* tailwind typo

* auto ack on mute

* don't change deref

* sqlx prep

* avoid renaming db column, keep acknowledged instead of acknowledged_global

* hide critical alert menu when not ee

* if workspace muted, also acknowledge global when workspace set

* removing save button for mute setting and improve ux/responsiveness

* fix: deployment callbacks have a concurrency limit of 1 on same path

* sqlx prep

* ee-repo ref

* z-[9999] for modal

---------

Co-authored-by: Guilhem <guilhemlemouel@gmail.com>
Co-authored-by: Ruben Fiszel <ruben@rubenfiszel.com>
Co-authored-by: Ruben Fiszel <ruben@windmill.dev>

.github/workflows/build_windows_worker_.yml

@@ -0,0 +1,58 @@
+name: Build windows executable for this branch
+
+on:
+  workflow_dispatch:
+
+env:
+  CARGO_INCREMENTAL: 0
+  SQLX_OFFLINE: true
+  DISABLE_EMBEDDING: true
+  RUST_LOG: info
+
+jobs:
+  cargo_build_windows:
+    runs-on: windows-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Read EE repo commit hash
+        shell: pwsh
+        run: |
+          $ee_repo_ref = Get-Content .\backend\ee-repo-ref.txt
+          echo "ee_repo_ref=$ee_repo_ref" | Out-File -FilePath $env:GITHUB_ENV -Append
+
+      - name: Checkout windmill-ee-private repository
+        uses: actions/checkout@v4
+        with:
+          repository: windmill-labs/windmill-ee-private
+          path: ./windmill-ee-private
+          ref: ${{ env.ee_repo_ref }}
+          token: ${{ secrets.WINDMILL_EE_PRIVATE_ACCESS }}
+          fetch-depth: 0
+
+      - name: Substitute EE code
+        shell: bash
+        run: |
+          ./backend/substitute_ee_code.sh --copy --dir ./windmill-ee-private
+
+      - name: Cargo build windows
+        timeout-minutes: 90
+        run: |
+          vcpkg.exe install openssl-windows:x64-windows
+          vcpkg.exe install openssl:x64-windows-static
+          vcpkg.exe integrate install
+          $env:VCPKGRS_DYNAMIC=1
+          $env:OPENSSL_DIR="${Env:VCPKG_INSTALLATION_ROOT}\installed\x64-windows-static"
+          mkdir frontend/build && cd backend
+          New-Item -Path . -Name "windmill-api/openapi-deref.yaml" -ItemType "File" -Force
+          cargo build --release --features=enterprise,stripe,embedding,parquet,prometheus,openidconnect,cloud,jemalloc,tantivy,deno_core
+
+      - name: Rename binary with corresponding architecture
+        run: |
+          Rename-Item -Path ".\backend\target\release\windmill.exe" -NewName "windmill-ee.exe"
+
+      - name: Upload artifact
+        uses: actions/upload-artifact@v3
+        with:
+          name: windmill-ee-binary
+          path: ./backend/target/release/windmill-ee.exe

.gitignore

@@ -6,4 +6,4 @@ frontend/src/routes/test.svelte
 CaddyfileRemoteMalo
 *.swp
 **/.idea/
-.direnv
+.direnv

CHANGELOG.md

@@ -1,5 +1,78 @@
 # Changelog
 
+## [1.424.0](https://github.com/windmill-labs/windmill/compare/v1.423.2...v1.424.0) (2024-11-14)
+
+
+### Features
+
+* allow setting password and login type from superadmin UI ([2a1bff3](https://github.com/windmill-labs/windmill/commit/2a1bff3160b65eadba399229b5f53950ec2c0d48))
+* **backend:** monitor minimal version of living workers ([#4704](https://github.com/windmill-labs/windmill/issues/4704)) ([50ff183](https://github.com/windmill-labs/windmill/commit/50ff183baeaa2a39c2b0188ee8dd6172b08ae801))
+* Support mistral anthropic for ai ([#4692](https://github.com/windmill-labs/windmill/issues/4692)) ([556b4a4](https://github.com/windmill-labs/windmill/commit/556b4a41a1d010bb8a1796f485d343c1e412d278))
+
+
+### Bug Fixes
+
+* add countCompletedJobs api ([2a78359](https://github.com/windmill-labs/windmill/commit/2a78359af7b8e4700634c2ad2745c1d14d2da4f7))
+* add queue_couts api ([10b6b1d](https://github.com/windmill-labs/windmill/commit/10b6b1dc04cb2b2d4ec5ceb26a5dbd2ac7bd1394))
+* autoscaling when count < min worker set to min_workers ([180bb82](https://github.com/windmill-labs/windmill/commit/180bb826436f9960def8c6cf3e9692714ffdf917))
+* deployment callbacks have a concurrency limit of 1 on same path ([c44fc4f](https://github.com/windmill-labs/windmill/commit/c44fc4f5ab8bcb5dffec08d5ed4ccd61feb1cb13))
+
+## [1.423.2](https://github.com/windmill-labs/windmill/compare/v1.423.1...v1.423.2) (2024-11-13)
+
+
+### Bug Fixes
+
+* fix intempestive expr type change in flow transform ([84bd06e](https://github.com/windmill-labs/windmill/commit/84bd06e90650e949c11e2b09447a7bad1ab60a95))
+
+## [1.423.1](https://github.com/windmill-labs/windmill/compare/v1.423.0...v1.423.1) (2024-11-12)
+
+
+### Bug Fixes
+
+* **autoscaling:** autoscaling thresholds to be >= and not > ([de00944](https://github.com/windmill-labs/windmill/commit/de00944f09699ff33a382f5a14f515ccf90b2454))
+
+## [1.423.0](https://github.com/windmill-labs/windmill/compare/v1.422.1...v1.423.0) (2024-11-12)
+
+
+### Features
+
+* s3 input available for public apps ([#4685](https://github.com/windmill-labs/windmill/issues/4685)) ([1671005](https://github.com/windmill-labs/windmill/commit/167100510032ab53cd609fb2c7629e67faceb093))
+
+
+### Bug Fixes
+
+* prevent underflow for autoscaling scalein ([5e9870a](https://github.com/windmill-labs/windmill/commit/5e9870a8a993032ec7d45c62e0023008da42c74a))
+* support multiple pip-extra-index-url with commas ([50861fc](https://github.com/windmill-labs/windmill/commit/50861fccc0cd86d4c76120c3306adf94f375330e))
+
+## [1.422.1](https://github.com/windmill-labs/windmill/compare/v1.422.0...v1.422.1) (2024-11-12)
+
+
+### Bug Fixes
+
+* fix password inputs ([e5e174a](https://github.com/windmill-labs/windmill/commit/e5e174ae9516f4c6b94ceb6e258b467f5c9a1f1a))
+
+## [1.422.0](https://github.com/windmill-labs/windmill/compare/v1.421.2...v1.422.0) (2024-11-11)
+
+
+### Features
+
+* expandable subflows in flows ([#4683](https://github.com/windmill-labs/windmill/issues/4683)) ([d44976f](https://github.com/windmill-labs/windmill/commit/d44976f35e45ade510d1ec220b5a1503e11f3db9))
+* **frontend:** critical alerts UI ([#4653](https://github.com/windmill-labs/windmill/issues/4653)) ([d9148ea](https://github.com/windmill-labs/windmill/commit/d9148eaa78680a93d81d71847a7df67e01f3c110))
+
+## [1.421.2](https://github.com/windmill-labs/windmill/compare/v1.421.1...v1.421.2) (2024-11-08)
+
+
+### Bug Fixes
+
+* **bash:** correctly propagate exit errors ([8bc9a02](https://github.com/windmill-labs/windmill/commit/8bc9a021a88391147c9170f56b5a0edddd55bc7d))
+
+## [1.421.1](https://github.com/windmill-labs/windmill/compare/v1.421.0...v1.421.1) (2024-11-08)
+
+
+### Bug Fixes
+
+* **python-client:** fix small break params of write_s3_file ([3ea12f1](https://github.com/windmill-labs/windmill/commit/3ea12f1821500e8b84549b892e3e1bb56a6ace4b))
+
 ## [1.421.0](https://github.com/windmill-labs/windmill/compare/v1.420.1...v1.421.0) (2024-11-07)
 
 

backend/.gitignore

@@ -2,7 +2,7 @@ target/
 .env
 oauth.json
 oauth2.json
-windmill-api/openapi-deref.yaml
 tracing.folded
 heaptrack*
 index/
+windmill-api/openapi-*.*

backend/.sqlx/query-00588a40dde5189ac1c61505f17acb0f4c244c60477427505bf5bd1b104d3bf9.json

@@ -0,0 +1,14 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE alerts SET acknowledged_workspace = true, acknowledged = true WHERE workspace_id = $1",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "00588a40dde5189ac1c61505f17acb0f4c244c60477427505bf5bd1b104d3bf9"
+}

backend/.sqlx/query-00ce4ed3ca0eac7cb6283b047353a64b9e78c4beb423f04baef9a53fbf87e9f9.json

@@ -1,14 +0,0 @@
-{
-  "db_name": "PostgreSQL",
-  "query": "INSERT INTO alerts (alert_type, message) VALUES ('recovered_critical_error', $1)",
-  "describe": {
-    "columns": [],
-    "parameters": {
-      "Left": [
-        "Text"
-      ]
-    },
-    "nullable": []
-  },
-  "hash": "00ce4ed3ca0eac7cb6283b047353a64b9e78c4beb423f04baef9a53fbf87e9f9"
-}

backend/.sqlx/query-0230bd64d2b6719c3536a106e18a68c7b43b3a9a9efa92b5517132e9c0d8a25b.json

@@ -1,12 +1,12 @@
 {
   "db_name": "PostgreSQL",
-  "query": "SELECT openai_resource_path, code_completion_enabled FROM workspace_settings WHERE workspace_id = $1",
+  "query": "SELECT ai_resource, code_completion_enabled FROM workspace_settings WHERE workspace_id = $1",
   "describe": {
     "columns": [
       {
         "ordinal": 0,
-        "name": "openai_resource_path",
-        "type_info": "Varchar"
+        "name": "ai_resource",
+        "type_info": "Jsonb"
       },
       {
         "ordinal": 1,
@@ -24,5 +24,5 @@
       false
     ]
   },
-  "hash": "ceb97024ebf1a1c00ea1ed4952f66b229ca4622c537cc252d8ed52b4d24270ee"
+  "hash": "0230bd64d2b6719c3536a106e18a68c7b43b3a9a9efa92b5517132e9c0d8a25b"
 }

backend/.sqlx/query-0331a81262e2d3c1bcfaeb64617b11eb68ab4599d64e5e5af639a9ac5d791fd0.json

@@ -1,6 +1,6 @@
 {
   "db_name": "PostgreSQL",
-  "query": "SELECT\n                -- slack_team_id, \n                -- slack_name, \n                -- slack_command_script, \n                -- CASE WHEN slack_email = 'missing@email.xyz' THEN NULL ELSE slack_email END AS slack_email,\n                auto_invite_domain IS NOT NULL AS \"auto_invite_enabled!\",\n                CASE WHEN auto_invite_operator IS TRUE THEN 'operator' ELSE 'developer' END AS \"auto_invite_as!\", \n                CASE WHEN auto_add IS TRUE THEN 'add' ELSE 'invite' END AS \"auto_invite_mode!\", \n                webhook, \n                deploy_to, \n                error_handler, \n                openai_resource_path, \n                code_completion_enabled, \n                error_handler_extra_args, \n                error_handler_muted_on_cancel, \n                large_file_storage, \n                git_sync,\n                default_app,\n                default_scripts,\n                workspace.name\n            FROM workspace_settings\n            LEFT JOIN workspace ON workspace.id = workspace_settings.workspace_id\n            WHERE workspace_id = $1",
+  "query": "SELECT\n                -- slack_team_id, \n                -- slack_name, \n                -- slack_command_script, \n                -- CASE WHEN slack_email = 'missing@email.xyz' THEN NULL ELSE slack_email END AS slack_email,\n                auto_invite_domain IS NOT NULL AS \"auto_invite_enabled!\",\n                CASE WHEN auto_invite_operator IS TRUE THEN 'operator' ELSE 'developer' END AS \"auto_invite_as!\", \n                CASE WHEN auto_add IS TRUE THEN 'add' ELSE 'invite' END AS \"auto_invite_mode!\", \n                webhook, \n                deploy_to, \n                error_handler, \n                ai_resource, \n                code_completion_enabled, \n                error_handler_extra_args, \n                error_handler_muted_on_cancel, \n                large_file_storage, \n                git_sync,\n                default_app,\n                default_scripts,\n                workspace.name\n            FROM workspace_settings\n            LEFT JOIN workspace ON workspace.id = workspace_settings.workspace_id\n            WHERE workspace_id = $1",
   "describe": {
     "columns": [
       {
@@ -35,8 +35,8 @@
       },
       {
         "ordinal": 6,
-        "name": "openai_resource_path",
-        "type_info": "Varchar"
+        "name": "ai_resource",
+        "type_info": "Jsonb"
       },
       {
         "ordinal": 7,
@@ -102,5 +102,5 @@
       false
     ]
   },
-  "hash": "188534f4b29f6461b1a6214d060f183c830b19a403ebb7b8be55a691675010c3"
+  "hash": "0331a81262e2d3c1bcfaeb64617b11eb68ab4599d64e5e5af639a9ac5d791fd0"
 }

backend/.sqlx/query-034583442e6f8ae38d6c4e4aac26f17c8d9d0e657f28276228fc90d3e22e1304.json

@@ -1,16 +0,0 @@
-{
-  "db_name": "PostgreSQL",
-  "query": "UPDATE workspace_settings SET openai_resource_path = $1, code_completion_enabled = $2 WHERE workspace_id = $3",
-  "describe": {
-    "columns": [],
-    "parameters": {
-      "Left": [
-        "Varchar",
-        "Bool",
-        "Text"
-      ]
-    },
-    "nullable": []
-  },
-  "hash": "034583442e6f8ae38d6c4e4aac26f17c8d9d0e657f28276228fc90d3e22e1304"
-}

backend/.sqlx/query-044e2b428ee6e2dd4543c87ad8835e239cf7567d18b8b3fa6608ea3a9d206ca7.json

@@ -0,0 +1,18 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "INSERT INTO alerts (alert_type, message, acknowledged, acknowledged_workspace, workspace_id, resource)\n        VALUES ('critical_error', $1, $2, $3, $4, $5)",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Text",
+        "Bool",
+        "Bool",
+        "Text",
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "044e2b428ee6e2dd4543c87ad8835e239cf7567d18b8b3fa6608ea3a9d206ca7"
+}

backend/.sqlx/query-0ee63ef2dd5c88edba2a1f56d31f29876724922f148dad7af35b36efbf70207a.json

@@ -1,14 +0,0 @@
-{
-  "db_name": "PostgreSQL",
-  "query": "DELETE FROM healthchecks WHERE check_type = $1",
-  "describe": {
-    "columns": [],
-    "parameters": {
-      "Left": [
-        "Text"
-      ]
-    },
-    "nullable": []
-  },
-  "hash": "0ee63ef2dd5c88edba2a1f56d31f29876724922f148dad7af35b36efbf70207a"
-}

backend/.sqlx/query-1610c79b8d238e3a35d795f34a2a0100b933b7f199d41cc2f554d57c811d55f3.json

@@ -0,0 +1,16 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE workspace_settings SET ai_resource = $1, code_completion_enabled = $2 WHERE workspace_id = $3",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Jsonb",
+        "Bool",
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "1610c79b8d238e3a35d795f34a2a0100b933b7f199d41cc2f554d57c811d55f3"
+}

backend/.sqlx/query-1730f39fd1793d45fbb41b21389c61296a3ff7489ae12f52a19f9543173ac597.json

@@ -65,8 +65,8 @@
       },
       {
         "ordinal": 12,
-        "name": "openai_resource_path",
-        "type_info": "Varchar"
+        "name": "ai_resource",
+        "type_info": "Jsonb"
       },
       {
         "ordinal": 13,
@@ -117,6 +117,11 @@
         "ordinal": 22,
         "name": "deploy_ui",
         "type_info": "Jsonb"
+      },
+      {
+        "ordinal": 23,
+        "name": "mute_critical_alerts",
+        "type_info": "Bool"
       }
     ],
     "parameters": {
@@ -147,6 +152,7 @@
       true,
       false,
       true,
+      true,
       true
     ]
   },

backend/.sqlx/query-1e7ce0c140410ae799f9c0c5772e4be4506bfd238c88f1b1c3ddf39b29071446.json

@@ -0,0 +1,29 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT\n                flow_status AS \"flow_status!: Json<Box<RawValue>>\",\n                raw_flow->'modules'->(flow_status->'step')::int AS \"module: Json<Box<RawValue>>\"\n            FROM queue WHERE id = $1 AND workspace_id = $2 LIMIT 1",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "flow_status!: Json<Box<RawValue>>",
+        "type_info": "Jsonb"
+      },
+      {
+        "ordinal": 1,
+        "name": "module: Json<Box<RawValue>>",
+        "type_info": "Jsonb"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Uuid",
+        "Text"
+      ]
+    },
+    "nullable": [
+      true,
+      null
+    ]
+  },
+  "hash": "1e7ce0c140410ae799f9c0c5772e4be4506bfd238c88f1b1c3ddf39b29071446"
+}

backend/.sqlx/query-286fa00c088df146c08c1934556f06a4243c66787ab8759a8045e60effd2fb77.json

@@ -1,6 +1,6 @@
 {
   "db_name": "PostgreSQL",
-  "query": "SELECT value\n            FROM resource\n            WHERE path = $1 AND workspace_id = $2",
+  "query": "SELECT value\n                FROM resource\n                WHERE path = $1 AND workspace_id = $2",
   "describe": {
     "columns": [
       {
@@ -19,5 +19,5 @@
       true
     ]
   },
-  "hash": "acdaa5151f8f7f37bb8c8c5a7d146789887e47db9695fc26b1dfaedd735e1e60"
+  "hash": "286fa00c088df146c08c1934556f06a4243c66787ab8759a8045e60effd2fb77"
 }

backend/.sqlx/query-28987898c7e4b172d466bf08f33c2da733e71f8a253b0122c17b9e021919809e.json

@@ -0,0 +1,18 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "INSERT INTO alerts (alert_type, message, acknowledged, acknowledged_workspace, workspace_id, resource)\n        VALUES ('recovered_critical_error', $1, $2, $3, $4, $5)",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Text",
+        "Bool",
+        "Bool",
+        "Text",
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "28987898c7e4b172d466bf08f33c2da733e71f8a253b0122c17b9e021919809e"
+}

backend/.sqlx/query-31dd23f6768052e486668172e15eba1a3f9b6a8e5678a7ee8e5456ff4405d6f9.json

@@ -0,0 +1,22 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT raw_flow->'failure_module' != 'null'::jsonb\n            FROM completed_job\n            WHERE id = $1",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "?column?",
+        "type_info": "Bool"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Uuid"
+      ]
+    },
+    "nullable": [
+      null
+    ]
+  },
+  "hash": "31dd23f6768052e486668172e15eba1a3f9b6a8e5678a7ee8e5456ff4405d6f9"
+}

backend/.sqlx/query-344b3a5d9683273a956b5156fed5ab9fdff1a7252198e4ae290d8c8f937ff177.json

@@ -0,0 +1,54 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT id, alert_type, message, created_at, COALESCE(acknowledged, false) AS acknowledged, workspace_id\n                 FROM alerts\n                 WHERE COALESCE(acknowledged, false) = $1\n                 ORDER BY created_at DESC\n                 LIMIT $2 OFFSET $3",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "id",
+        "type_info": "Int4"
+      },
+      {
+        "ordinal": 1,
+        "name": "alert_type",
+        "type_info": "Varchar"
+      },
+      {
+        "ordinal": 2,
+        "name": "message",
+        "type_info": "Text"
+      },
+      {
+        "ordinal": 3,
+        "name": "created_at",
+        "type_info": "Timestamptz"
+      },
+      {
+        "ordinal": 4,
+        "name": "acknowledged",
+        "type_info": "Bool"
+      },
+      {
+        "ordinal": 5,
+        "name": "workspace_id",
+        "type_info": "Text"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Bool",
+        "Int8",
+        "Int8"
+      ]
+    },
+    "nullable": [
+      false,
+      false,
+      false,
+      false,
+      null,
+      true
+    ]
+  },
+  "hash": "344b3a5d9683273a956b5156fed5ab9fdff1a7252198e4ae290d8c8f937ff177"
+}

backend/.sqlx/query-349396e8fdd96d45875110bc06767e6eb876792ec8f83e9b03c2fb46bb12e0b9.json

@@ -0,0 +1,15 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE password SET login_type = $1 WHERE email = $2",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Varchar",
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "349396e8fdd96d45875110bc06767e6eb876792ec8f83e9b03c2fb46bb12e0b9"
+}

backend/.sqlx/query-4a3a8207627418ba7b7eabaccd7ec72884f2a1eacf0ab0e0c6ccbf56c654c14a.json

@@ -0,0 +1,14 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE alerts SET acknowledged = true, acknowledged_workspace = true WHERE resource = $1 AND alert_type = 'critical_error'",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "4a3a8207627418ba7b7eabaccd7ec72884f2a1eacf0ab0e0c6ccbf56c654c14a"
+}

backend/.sqlx/query-4c0067c2135a259aea5cc2db60f7375a9a33671be8ef406427d90f67a98c9c9f.json

@@ -1,14 +0,0 @@
-{
-  "db_name": "PostgreSQL",
-  "query": "INSERT INTO alerts (alert_type, message) VALUES ('critical_error', $1)",
-  "describe": {
-    "columns": [],
-    "parameters": {
-      "Left": [
-        "Text"
-      ]
-    },
-    "nullable": []
-  },
-  "hash": "4c0067c2135a259aea5cc2db60f7375a9a33671be8ef406427d90f67a98c9c9f"
-}

backend/.sqlx/query-4d30c5a2894d655741d167f5589f5816583f0bae78a0dd3270dedb377dfa944a.json

@@ -0,0 +1,53 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT id, alert_type, message, created_at, COALESCE(acknowledged, false) AS acknowledged, workspace_id\n                 FROM alerts\n                 ORDER BY created_at DESC\n                 LIMIT $1 OFFSET $2",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "id",
+        "type_info": "Int4"
+      },
+      {
+        "ordinal": 1,
+        "name": "alert_type",
+        "type_info": "Varchar"
+      },
+      {
+        "ordinal": 2,
+        "name": "message",
+        "type_info": "Text"
+      },
+      {
+        "ordinal": 3,
+        "name": "created_at",
+        "type_info": "Timestamptz"
+      },
+      {
+        "ordinal": 4,
+        "name": "acknowledged",
+        "type_info": "Bool"
+      },
+      {
+        "ordinal": 5,
+        "name": "workspace_id",
+        "type_info": "Text"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Int8",
+        "Int8"
+      ]
+    },
+    "nullable": [
+      false,
+      false,
+      false,
+      false,
+      null,
+      true
+    ]
+  },
+  "hash": "4d30c5a2894d655741d167f5589f5816583f0bae78a0dd3270dedb377dfa944a"
+}

backend/.sqlx/query-55cb03040bc2a8c53dd7fbb42bbdcc40f463cbc52d94ed9315cf9a547d4c89f2.json

@@ -65,8 +65,8 @@
       },
       {
         "ordinal": 12,
-        "name": "openai_resource_path",
-        "type_info": "Varchar"
+        "name": "ai_resource",
+        "type_info": "Jsonb"
       },
       {
         "ordinal": 13,
@@ -117,6 +117,11 @@
         "ordinal": 22,
         "name": "deploy_ui",
         "type_info": "Jsonb"
+      },
+      {
+        "ordinal": 23,
+        "name": "mute_critical_alerts",
+        "type_info": "Bool"
       }
     ],
     "parameters": {
@@ -147,6 +152,7 @@
       true,
       false,
       true,
+      true,
       true
     ]
   },

backend/.sqlx/query-6268eabd561502a44e273d6391102278974623f7a7bcad6891d7abadf4d3ea03.json

@@ -1,15 +0,0 @@
-{
-  "db_name": "PostgreSQL",
-  "query": "UPDATE workspace_settings SET openai_resource_path = NULL, code_completion_enabled = $1 WHERE workspace_id = $2",
-  "describe": {
-    "columns": [],
-    "parameters": {
-      "Left": [
-        "Bool",
-        "Text"
-      ]
-    },
-    "nullable": []
-  },
-  "hash": "6268eabd561502a44e273d6391102278974623f7a7bcad6891d7abadf4d3ea03"
-}

backend/.sqlx/query-65da41c7ded54cdee8d33211561c068b72294cc99ff44ed0a13179df508ebc6a.json

@@ -0,0 +1,15 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE alerts\n         SET\n           acknowledged = true,\n           acknowledged_workspace = CASE\n             WHEN $2::text IS NOT NULL AND workspace_id = $2 THEN true\n             ELSE acknowledged_workspace\n           END\n         WHERE id = $1",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Int4",
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "65da41c7ded54cdee8d33211561c068b72294cc99ff44ed0a13179df508ebc6a"
+}

backend/.sqlx/query-661f472ff3860983322162420457f5033b9c9afc344d9c3e385ba20a3ad2197a.json

@@ -5,7 +5,7 @@
     "columns": [
       {
         "ordinal": 0,
-        "name": "bool",
+        "name": "?column?",
         "type_info": "Bool"
       }
     ],

backend/.sqlx/query-6940ddc5ee8d2a1048213d46f52d964b199604ba66dedbe9f89cea727d726a2d.json

@@ -0,0 +1,15 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE workspace_settings SET ai_resource = NULL, code_completion_enabled = $1 WHERE workspace_id = $2",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Bool",
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "6940ddc5ee8d2a1048213d46f52d964b199604ba66dedbe9f89cea727d726a2d"
+}

backend/.sqlx/query-777e7084edf9a5d14f299ff479ae43c8ead47a6a531d4b031a1342c9b2f16506.json

@@ -0,0 +1,54 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT id, alert_type, message, created_at, COALESCE(acknowledged_workspace, false) AS acknowledged, workspace_id\n                 FROM alerts\n                 WHERE workspace_id = $1\n                 ORDER BY created_at DESC\n                 LIMIT $2 OFFSET $3",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "id",
+        "type_info": "Int4"
+      },
+      {
+        "ordinal": 1,
+        "name": "alert_type",
+        "type_info": "Varchar"
+      },
+      {
+        "ordinal": 2,
+        "name": "message",
+        "type_info": "Text"
+      },
+      {
+        "ordinal": 3,
+        "name": "created_at",
+        "type_info": "Timestamptz"
+      },
+      {
+        "ordinal": 4,
+        "name": "acknowledged",
+        "type_info": "Bool"
+      },
+      {
+        "ordinal": 5,
+        "name": "workspace_id",
+        "type_info": "Text"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Text",
+        "Int8",
+        "Int8"
+      ]
+    },
+    "nullable": [
+      false,
+      false,
+      false,
+      false,
+      null,
+      true
+    ]
+  },
+  "hash": "777e7084edf9a5d14f299ff479ae43c8ead47a6a531d4b031a1342c9b2f16506"
+}

backend/.sqlx/query-7c32176755c6ea2b6ae531860d436caae3fa256fc0803749ec5107632669adb3.json

@@ -0,0 +1,14 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE alerts \n         SET\n           acknowledged = true,\n           acknowledged_workspace = CASE\n             WHEN $1::text IS NOT NULL THEN true\n             ELSE acknowledged_workspace\n           END\n         WHERE ($1::text IS NOT NULL AND workspace_id = $1)\n            OR ($1::text IS NULL)",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "7c32176755c6ea2b6ae531860d436caae3fa256fc0803749ec5107632669adb3"
+}

backend/.sqlx/query-7c5a29de07cbe42326a15d4d7fd9714c20b767508e27368be6de38d7b18a3a4d.json

@@ -0,0 +1,55 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT id, alert_type, message, created_at, COALESCE(acknowledged_workspace, false) AS acknowledged, workspace_id\n                 FROM alerts\n                 WHERE workspace_id = $1 AND COALESCE(acknowledged_workspace, false) = $2\n                 ORDER BY created_at DESC\n                 LIMIT $3 OFFSET $4",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "id",
+        "type_info": "Int4"
+      },
+      {
+        "ordinal": 1,
+        "name": "alert_type",
+        "type_info": "Varchar"
+      },
+      {
+        "ordinal": 2,
+        "name": "message",
+        "type_info": "Text"
+      },
+      {
+        "ordinal": 3,
+        "name": "created_at",
+        "type_info": "Timestamptz"
+      },
+      {
+        "ordinal": 4,
+        "name": "acknowledged",
+        "type_info": "Bool"
+      },
+      {
+        "ordinal": 5,
+        "name": "workspace_id",
+        "type_info": "Text"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Text",
+        "Bool",
+        "Int8",
+        "Int8"
+      ]
+    },
+    "nullable": [
+      false,
+      false,
+      false,
+      false,
+      null,
+      true
+    ]
+  },
+  "hash": "7c5a29de07cbe42326a15d4d7fd9714c20b767508e27368be6de38d7b18a3a4d"
+}

backend/.sqlx/query-a59fae29ebcc9aa53308b777ead3d2b652618ac454f139db738f4499cb4eb079.json

@@ -0,0 +1,12 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE alerts SET acknowledged = true",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": []
+    },
+    "nullable": []
+  },
+  "hash": "a59fae29ebcc9aa53308b777ead3d2b652618ac454f139db738f4499cb4eb079"
+}

backend/.sqlx/query-a69ba7471d1a5faa145bebbd17d43e6dbe02e9e92ebbc31a50c99c3a04719284.json

@@ -1,6 +1,6 @@
 {
   "db_name": "PostgreSQL",
-  "query": "WITH uuid_table as (\n            select gen_random_uuid() as uuid from generate_series(1, $11)\n        )\n        INSERT INTO queue \n            (id, script_hash, script_path, job_kind, language, args, tag, created_by, permissioned_as, email, scheduled_for, workspace_id, concurrent_limit, concurrency_time_window_s, timeout)\n            (SELECT uuid, $1, $2, $3, $4, ('{ \"uuid\": \"' || uuid || '\" }')::jsonb, $5, $6, $7, $8, $9, $10, $12, $13, $14 FROM uuid_table) \n        RETURNING id",
+  "query": "WITH uuid_table as (\n            select gen_random_uuid() as uuid from generate_series(1, $11)\n        )\n        INSERT INTO queue \n            (id, script_hash, script_path, job_kind, language, args, tag, created_by, permissioned_as, email, scheduled_for, workspace_id, concurrent_limit, concurrency_time_window_s, timeout, raw_code, raw_lock, raw_flow, flow_status)\n            (SELECT uuid, $1, $2, $3, $4, ('{ \"uuid\": \"' || uuid || '\" }')::jsonb, $5, $6, $7, $8, $9, $10, $12, $13, $14, $15, $16, $17, $18 FROM uuid_table) \n        RETURNING id",
   "describe": {
     "columns": [
       {
@@ -72,12 +72,16 @@
         "Int4",
         "Int4",
         "Int4",
-        "Int4"
+        "Int4",
+        "Text",
+        "Text",
+        "Jsonb",
+        "Jsonb"
       ]
     },
     "nullable": [
       false
     ]
   },
-  "hash": "0a686ca61444d7ad7484071727aa039a6ea6697e5a49a633b767c052aa3e0a18"
+  "hash": "a69ba7471d1a5faa145bebbd17d43e6dbe02e9e92ebbc31a50c99c3a04719284"
 }

backend/.sqlx/query-aa6907b7266ee437dfbd77a9bf6c047fda88f3e72a0d10323a5e4020cf857c42.json

@@ -0,0 +1,22 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT raw_flow->'failure_module' != 'null'::jsonb\n            FROM queue\n            WHERE id = $1",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "?column?",
+        "type_info": "Bool"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Uuid"
+      ]
+    },
+    "nullable": [
+      null
+    ]
+  },
+  "hash": "aa6907b7266ee437dfbd77a9bf6c047fda88f3e72a0d10323a5e4020cf857c42"
+}

backend/.sqlx/query-ad03e5acf10ef94abc37cb9f56b1775c67f075c8bc83458e8be2e242347218d6.json

@@ -0,0 +1,22 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT wm_version FROM worker_ping WHERE wm_version != $1 AND ping_at > now() - interval '5 minutes'",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "wm_version",
+        "type_info": "Varchar"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Text"
+      ]
+    },
+    "nullable": [
+      false
+    ]
+  },
+  "hash": "ad03e5acf10ef94abc37cb9f56b1775c67f075c8bc83458e8be2e242347218d6"
+}

backend/.sqlx/query-ae543dfa106fa6ad4e9bf45cda1110d4702a80f455c63af6fcbd7cf45bbc4f7a.json

@@ -0,0 +1,23 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT flow_version.value AS \"value: sqlx::types::Json<Box<RawValue>>\" FROM flow \n                    LEFT JOIN flow_version\n                        ON flow_version.id = flow.versions[array_upper(flow.versions, 1)]\n                    WHERE flow.path = $1 AND flow.workspace_id = $2",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "value: sqlx::types::Json<Box<RawValue>>",
+        "type_info": "Jsonb"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Text",
+        "Text"
+      ]
+    },
+    "nullable": [
+      false
+    ]
+  },
+  "hash": "ae543dfa106fa6ad4e9bf45cda1110d4702a80f455c63af6fcbd7cf45bbc4f7a"
+}

backend/.sqlx/query-ceda377fa534656ac12a7f41db77db1f054ec751855c8db7352b0e9c20b63ef8.json

@@ -0,0 +1,22 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "SELECT ai_resource FROM workspace_settings WHERE workspace_id = $1",
+  "describe": {
+    "columns": [
+      {
+        "ordinal": 0,
+        "name": "ai_resource",
+        "type_info": "Jsonb"
+      }
+    ],
+    "parameters": {
+      "Left": [
+        "Text"
+      ]
+    },
+    "nullable": [
+      true
+    ]
+  },
+  "hash": "ceda377fa534656ac12a7f41db77db1f054ec751855c8db7352b0e9c20b63ef8"
+}

backend/.sqlx/query-de1abe57b6aa61155f747a3bcb98359f70eade6654b5a884915f07f3ef3fe15e.json

@@ -1,6 +1,6 @@
 {
   "db_name": "PostgreSQL",
-  "query": "SELECT raw_flow->'modules'->($1)->'value'->>'type' = 'flow' FROM queue WHERE id = $2",
+  "query": "SELECT raw_flow->'modules'->($1)::text->'value'->>'type' = 'flow' FROM queue WHERE id = $2 LIMIT 1",
   "describe": {
     "columns": [
       {
@@ -19,5 +19,5 @@
       null
     ]
   },
-  "hash": "3e539fef054ad31bc1736e27276087775a721a6ee7ae35b03fd4ce3563ea3838"
+  "hash": "de1abe57b6aa61155f747a3bcb98359f70eade6654b5a884915f07f3ef3fe15e"
 }

backend/.sqlx/query-e05dbe046e846c092a96b6b0a9d872c721169d418586be2d15cee6b929049098.json

@@ -1,12 +1,12 @@
 {
   "db_name": "PostgreSQL",
-  "query": "SELECT openai_resource_path FROM workspace_settings WHERE workspace_id = $1",
+  "query": "SELECT mute_critical_alerts FROM workspace_settings WHERE workspace_id = $1",
   "describe": {
     "columns": [
       {
         "ordinal": 0,
-        "name": "openai_resource_path",
-        "type_info": "Varchar"
+        "name": "mute_critical_alerts",
+        "type_info": "Bool"
       }
     ],
     "parameters": {
@@ -18,5 +18,5 @@
       true
     ]
   },
-  "hash": "f8b689ec3e09f14dc02ac9e9f98d252a14d3a2acfadcd9cd6ba27ca9799f4706"
+  "hash": "e05dbe046e846c092a96b6b0a9d872c721169d418586be2d15cee6b929049098"
 }

backend/.sqlx/query-f22168826350797e88153b65a5b1e906a7868f34c062f154b2ee4f24c57aa5c3.json

@@ -0,0 +1,15 @@
+{
+  "db_name": "PostgreSQL",
+  "query": "UPDATE workspace_settings SET mute_critical_alerts = $1 WHERE workspace_id = $2",
+  "describe": {
+    "columns": [],
+    "parameters": {
+      "Left": [
+        "Bool",
+        "Text"
+      ]
+    },
+    "nullable": []
+  },
+  "hash": "f22168826350797e88153b65a5b1e906a7868f34c062f154b2ee4f24c57aa5c3"
+}

backend/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "windmill"
-version = "1.421.0"
+version = "1.424.0"
 authors.workspace = true
 edition.workspace = true
 
@@ -29,7 +29,7 @@ members = [
 ]
 
 [workspace.package]
-version = "1.421.0"
+version = "1.424.0"
 authors = ["Ruben Fiszel <ruben@windmill.dev>"]
 edition = "2021"
 
@@ -180,6 +180,7 @@ tokio-util = { version = "^0", features = ["io"] }
 json-pointer = "^0"
 itertools = "^0"
 regex = "^1"
+semver = "^1"
 
 deno_fetch = "0.195.0"
 deno_tls = "0.158.0"

backend/ee-repo-ref.txt

@@ -1 +1 @@
-6534b0f31fb4a75dd706fca2ce91e37e77e4ad02
+51dcbf93b0d127af9f33fa346cc63fcd2475d4fa

backend/migrations/20241105133531_alter_openai_resource_path_column_in_workspace_settings.down.sql

@@ -0,0 +1,7 @@
+-- Add down migration script here
+ALTER TABLE workspace_settings
+ALTER COLUMN ai_resource TYPE text
+USING ai_resource->>'path';
+
+ALTER TABLE workspace_settings
+RENAME COLUMN ai_resource to openai_resource_path;

backend/migrations/20241105133531_alter_openai_resource_path_column_in_workspace_settings.up.sql

@@ -0,0 +1,7 @@
+-- Add up migration script here
+ALTER TABLE workspace_settings
+ALTER COLUMN openai_resource_path TYPE jsonb
+USING jsonb_build_object('provider', 'openai', 'path', openai_resource_path);
+
+ALTER TABLE workspace_settings
+RENAME COLUMN openai_resource_path TO ai_resource;

backend/migrations/20241105225151_alert_add_acknowledged.down.sql

@@ -0,0 +1 @@
+ALTER TABLE alerts DROP COLUMN acknowledged;

backend/migrations/20241105225151_alert_add_acknowledged.up.sql

@@ -0,0 +1,9 @@
+-- Step 1: Add the new column 'acknowledged' to the 'alerts' table
+ALTER TABLE alerts
+ADD COLUMN acknowledged BOOLEAN DEFAULT NULL;
+
+-- Step 2: Update all existing rows to set 'acknowledged' to true
+-- we don't want to pop up notifications to all users after migrations
+-- but only show new alerts from the point of the upgrade
+UPDATE alerts
+SET acknowledged = TRUE;

backend/migrations/20241108201445_alert_add_workspace_id.down.sql

@@ -0,0 +1,4 @@
+ALTER TABLE alerts DROP COLUMN workspace_id;
+ALTER TABLE alerts DROP COLUMN acknowledged_workspace;
+ALTER TABLE alerts DROP COLUMN resource;
+ALTER TABLE workspace_settings DROP COLUMN mute_critical_alerts;

backend/migrations/20241108201445_alert_add_workspace_id.up.sql

@@ -0,0 +1,4 @@
+ALTER TABLE alerts ADD COLUMN workspace_id TEXT DEFAULT NULL;
+ALTER TABLE alerts ADD COLUMN acknowledged_workspace BOOL DEFAULT NULL;
+ALTER TABLE alerts ADD COLUMN resource TEXT DEFAULT NULL;
+ALTER TABLE workspace_settings ADD COLUMN mute_critical_alerts BOOL DEFAULT NULL;

backend/parsers/windmill-parser-py-imports/src/lib.rs

@@ -62,6 +62,7 @@ static PYTHON_IMPORTS_REPLACEMENT: phf::Map<&'static str, &'static str> = phf_ma
     "lokalise" => "python-lokalise-api",
     "msgraph" => "msgraph-sdk",
     "pythonjsonlogger" => "python-json-logger",
+    "socks" => "PySocks",
 };
 
 fn replace_import(x: String) -> String {

backend/src/main.rs

@@ -7,7 +7,6 @@
  */
 
 use anyhow::Context;
-use git_version::git_version;
 use monitor::{
     reload_timeout_wait_result_setting, send_current_log_file_to_object_store,
     send_logs_to_object_store,
@@ -31,18 +30,19 @@ use windmill_common::ee::{maybe_renew_license_key_on_start, LICENSE_KEY_ID, LICE
 
 use windmill_common::{
     global_settings::{
-        BASE_URL_SETTING, BUNFIG_INSTALL_SCOPES_SETTING, CRITICAL_ERROR_CHANNELS_SETTING,
-        CUSTOM_TAGS_SETTING, DEFAULT_TAGS_PER_WORKSPACE_SETTING, DEFAULT_TAGS_WORKSPACES_SETTING,
-        ENV_SETTINGS, EXPOSE_DEBUG_METRICS_SETTING, EXPOSE_METRICS_SETTING,
-        EXTRA_PIP_INDEX_URL_SETTING, HUB_BASE_URL_SETTING, JOB_DEFAULT_TIMEOUT_SECS_SETTING,
-        JWT_SECRET_SETTING, KEEP_JOB_DIR_SETTING, LICENSE_KEY_SETTING, NPM_CONFIG_REGISTRY_SETTING,
-        OAUTH_SETTING, PIP_INDEX_URL_SETTING, REQUEST_SIZE_LIMIT_SETTING,
-        REQUIRE_PREEXISTING_USER_FOR_OAUTH_SETTING, RETENTION_PERIOD_SECS_SETTING,
-        SAML_METADATA_SETTING, SCIM_TOKEN_SETTING, SMTP_SETTING, TIMEOUT_WAIT_RESULT_SETTING,
+        BASE_URL_SETTING, BUNFIG_INSTALL_SCOPES_SETTING, CRITICAL_ALERT_MUTE_UI_SETTING,
+        CRITICAL_ERROR_CHANNELS_SETTING, CUSTOM_TAGS_SETTING, DEFAULT_TAGS_PER_WORKSPACE_SETTING,
+        DEFAULT_TAGS_WORKSPACES_SETTING, ENV_SETTINGS, EXPOSE_DEBUG_METRICS_SETTING,
+        EXPOSE_METRICS_SETTING, EXTRA_PIP_INDEX_URL_SETTING, HUB_BASE_URL_SETTING,
+        JOB_DEFAULT_TIMEOUT_SECS_SETTING, JWT_SECRET_SETTING, KEEP_JOB_DIR_SETTING,
+        LICENSE_KEY_SETTING, NPM_CONFIG_REGISTRY_SETTING, OAUTH_SETTING, PIP_INDEX_URL_SETTING,
+        REQUEST_SIZE_LIMIT_SETTING, REQUIRE_PREEXISTING_USER_FOR_OAUTH_SETTING,
+        RETENTION_PERIOD_SECS_SETTING, SAML_METADATA_SETTING, SCIM_TOKEN_SETTING, SMTP_SETTING,
+        TIMEOUT_WAIT_RESULT_SETTING,
     },
     scripts::ScriptLang,
     stats_ee::schedule_stats,
-    utils::{hostname, rd_string, Mode},
+    utils::{hostname, rd_string, Mode, GIT_VERSION},
     worker::{reload_custom_tags_setting, HUB_CACHE_DIR, TMP_DIR, WORKER_GROUP},
     DB, METRICS_ENABLED,
 };
@@ -67,16 +67,17 @@ use windmill_worker::{
     get_hub_script_content_and_requirements, BUN_BUNDLE_CACHE_DIR, BUN_CACHE_DIR,
     BUN_DEPSTAR_CACHE_DIR, DENO_CACHE_DIR, DENO_CACHE_DIR_DEPS, DENO_CACHE_DIR_NPM,
     GO_BIN_CACHE_DIR, GO_CACHE_DIR, LOCK_CACHE_DIR, PIP_CACHE_DIR, POWERSHELL_CACHE_DIR,
-    RUST_CACHE_DIR, TAR_PIP_CACHE_DIR, TMP_LOGS_DIR, UV_CACHE_DIR,
+    PY311_CACHE_DIR, RUST_CACHE_DIR, TAR_PIP_CACHE_DIR, TMP_LOGS_DIR, UV_CACHE_DIR,
 };
 
 use crate::monitor::{
     initial_load, load_keep_job_dir, load_metrics_debug_enabled, load_require_preexisting_user,
     load_tag_per_workspace_enabled, load_tag_per_workspace_workspaces, monitor_db, monitor_pool,
     reload_base_url_setting, reload_bunfig_install_scopes_setting,
-    reload_critical_error_channels_setting, reload_extra_pip_index_url_setting,
-    reload_hub_base_url_setting, reload_job_default_timeout_setting, reload_jwt_secret_setting,
-    reload_license_key, reload_npm_config_registry_setting, reload_pip_index_url_setting,
+    reload_critical_alert_mute_ui_setting, reload_critical_error_channels_setting,
+    reload_extra_pip_index_url_setting, reload_hub_base_url_setting,
+    reload_job_default_timeout_setting, reload_jwt_secret_setting, reload_license_key,
+    reload_npm_config_registry_setting, reload_pip_index_url_setting,
     reload_retention_period_setting, reload_scim_token_setting, reload_smtp_config,
     reload_worker_config,
 };
@@ -84,7 +85,6 @@ use crate::monitor::{
 #[cfg(feature = "parquet")]
 use crate::monitor::reload_s3_cache_setting;
 
-const GIT_VERSION: &str = git_version!(args = ["--tag", "--always"], fallback = "unknown-version");
 const DEFAULT_NUM_WORKERS: usize = 1;
 const DEFAULT_PORT: u16 = 8000;
 const DEFAULT_SERVER_BIND_ADDR: Ipv4Addr = Ipv4Addr::new(0, 0, 0, 0);
@@ -817,6 +817,12 @@ Windmill Community Edition {GIT_VERSION}
                                                         tracing::error!(error = %e, "Could not reload jwt secret setting");
                                                     }
                                                 },
+                                                CRITICAL_ALERT_MUTE_UI_SETTING => {
+                                                    tracing::info!("Critical alert UI setting changed");
+                                                    if let Err(e) = reload_critical_alert_mute_ui_setting(&db).await {
+                                                        tracing::error!(error = %e, "Could not reload critical alert UI setting");
+                                                    }
+                                                },
                                                 a @_ => {
                                                     tracing::info!("Unrecognized Global Setting Change Payload: {:?}", a);
                                                 }
@@ -991,13 +997,14 @@ pub async fn run_workers<R: rsmq_async::RsmqConnection + Send + Sync + Clone + '
     for x in [
         LOCK_CACHE_DIR,
         TMP_LOGS_DIR,
-        PIP_CACHE_DIR,
         UV_CACHE_DIR,
         TAR_PIP_CACHE_DIR,
         DENO_CACHE_DIR,
         DENO_CACHE_DIR_DEPS,
         DENO_CACHE_DIR_NPM,
         BUN_CACHE_DIR,
+        PY311_CACHE_DIR,
+        PIP_CACHE_DIR,
         BUN_DEPSTAR_CACHE_DIR,
         BUN_BUNDLE_CACHE_DIR,
         GO_CACHE_DIR,

backend/src/monitor.rs

@@ -34,12 +34,12 @@ use windmill_common::{
     error,
     flow_status::FlowStatusModule,
     global_settings::{
-        BASE_URL_SETTING, BUNFIG_INSTALL_SCOPES_SETTING, CRITICAL_ERROR_CHANNELS_SETTING,
-        DEFAULT_TAGS_PER_WORKSPACE_SETTING, DEFAULT_TAGS_WORKSPACES_SETTING,
-        EXPOSE_DEBUG_METRICS_SETTING, EXPOSE_METRICS_SETTING, EXTRA_PIP_INDEX_URL_SETTING,
-        HUB_BASE_URL_SETTING, JOB_DEFAULT_TIMEOUT_SECS_SETTING, JWT_SECRET_SETTING,
-        KEEP_JOB_DIR_SETTING, LICENSE_KEY_SETTING, NPM_CONFIG_REGISTRY_SETTING, OAUTH_SETTING,
-        PIP_INDEX_URL_SETTING, REQUEST_SIZE_LIMIT_SETTING,
+        BASE_URL_SETTING, BUNFIG_INSTALL_SCOPES_SETTING, CRITICAL_ALERT_MUTE_UI_SETTING,
+        CRITICAL_ERROR_CHANNELS_SETTING, DEFAULT_TAGS_PER_WORKSPACE_SETTING,
+        DEFAULT_TAGS_WORKSPACES_SETTING, EXPOSE_DEBUG_METRICS_SETTING, EXPOSE_METRICS_SETTING,
+        EXTRA_PIP_INDEX_URL_SETTING, HUB_BASE_URL_SETTING, JOB_DEFAULT_TIMEOUT_SECS_SETTING,
+        JWT_SECRET_SETTING, KEEP_JOB_DIR_SETTING, LICENSE_KEY_SETTING, NPM_CONFIG_REGISTRY_SETTING,
+        OAUTH_SETTING, PIP_INDEX_URL_SETTING, REQUEST_SIZE_LIMIT_SETTING,
         REQUIRE_PREEXISTING_USER_FOR_OAUTH_SETTING, RETENTION_PERIOD_SECS_SETTING,
         SAML_METADATA_SETTING, SCIM_TOKEN_SETTING, TIMEOUT_WAIT_RESULT_SETTING,
     },
@@ -51,11 +51,11 @@ use windmill_common::{
     utils::{now_from_db, rd_string, report_critical_error, Mode},
     worker::{
         load_worker_config, make_pull_query, make_suspended_pull_query, reload_custom_tags_setting,
-        DEFAULT_TAGS_PER_WORKSPACE, DEFAULT_TAGS_WORKSPACES, SMTP_CONFIG, WORKER_CONFIG,
-        WORKER_GROUP,
+        update_min_version, DEFAULT_TAGS_PER_WORKSPACE, DEFAULT_TAGS_WORKSPACES, SMTP_CONFIG,
+        WORKER_CONFIG, WORKER_GROUP,
     },
-    BASE_URL, CRITICAL_ERROR_CHANNELS, DB, DEFAULT_HUB_BASE_URL, HUB_BASE_URL, JOB_RETENTION_SECS,
-    METRICS_DEBUG_ENABLED, METRICS_ENABLED,
+    BASE_URL, CRITICAL_ALERT_MUTE_UI_ENABLED, CRITICAL_ERROR_CHANNELS, DB, DEFAULT_HUB_BASE_URL,
+    HUB_BASE_URL, JOB_RETENTION_SECS, METRICS_DEBUG_ENABLED, METRICS_ENABLED,
 };
 use windmill_queue::cancel_job;
 use windmill_worker::{
@@ -131,6 +131,10 @@ pub async fn initial_load(
         tracing::error!("Error loading expose debug metrics: {e:#}");
     }
 
+    if let Err(e) = reload_critical_alert_mute_ui_setting(db).await {
+        tracing::error!("Error loading critical alert mute ui setting: {e:#}");
+    }
+
     if let Err(e) = load_tag_per_workspace_enabled(db).await {
         tracing::error!("Error loading default tag per workpsace: {e:#}");
     }
@@ -226,6 +230,24 @@ pub async fn load_tag_per_workspace_workspaces(db: &DB) -> error::Result<()> {
     Ok(())
 }
 
+pub async fn reload_critical_alert_mute_ui_setting(db: &DB) -> error::Result<()> {
+    if let Ok(Some(serde_json::Value::Bool(t))) =
+        load_value_from_global_settings(db, CRITICAL_ALERT_MUTE_UI_SETTING).await
+    {
+        CRITICAL_ALERT_MUTE_UI_ENABLED.store(t, Ordering::Relaxed);
+
+        if t {
+            if let Err(e) = sqlx::query!("UPDATE alerts SET acknowledged = true")
+                .execute(db)
+                .await
+            {
+                tracing::error!("Error updating alerts: {}", e.to_string());
+            }
+        }
+    }
+    Ok(())
+}
+
 pub async fn load_metrics_debug_enabled(db: &DB) -> error::Result<()> {
     let metrics_enabled = load_value_from_global_settings(db, EXPOSE_DEBUG_METRICS_SETTING).await;
     match metrics_enabled {
@@ -1065,6 +1087,10 @@ pub async fn monitor_db(
         }
     };
 
+    let update_min_worker_version_f = async {
+        update_min_version(db).await;
+    };
+
     join!(
         expired_items_f,
         zombie_jobs_f,
@@ -1073,6 +1099,7 @@ pub async fn monitor_db(
         worker_groups_alerts_f,
         jobs_waiting_alerts_f,
         apply_autoscaling_f,
+        update_min_worker_version_f,
     );
 }
 
@@ -1303,7 +1330,7 @@ async fn handle_zombie_jobs<R: rsmq_async::RsmqConnection + Send + Sync + Clone>
                 ON CONFLICT (job_id) DO UPDATE SET logs = job_logs.logs || '\nRestarted job after not receiving job''s ping for too long the ' || now() || '\n\n' WHERE job_logs.job_id = $1", r.id)
                 .execute(db).await;
             tracing::error!(error_message);
-            report_critical_error(error_message, db.clone()).await;
+            report_critical_error(error_message, db.clone(), Some(&r.workspace_id), None).await;
         }
     }
 
@@ -1417,7 +1444,7 @@ async fn handle_zombie_flows(
                 flow.id, flow.workspace_id
             );
             tracing::error!(error_message);
-            report_critical_error(error_message, db.clone()).await;
+            report_critical_error(error_message, db.clone(), Some(&flow.workspace_id), None).await;
             // if the flow hasn't started and is a zombie, we can simply restart it
             sqlx::query!(
                 "UPDATE queue SET running = false, started_at = null WHERE id = $1 AND canceled = false",
@@ -1437,7 +1464,7 @@ async fn handle_zombie_flows(
                     format!("Flow {id} was cancelled because it")
                 }
             );
-            report_critical_error(reason.clone(), db.clone()).await;
+            report_critical_error(reason.clone(), db.clone(), Some(&flow.workspace_id), None).await;
             cancel_zombie_flow_job(db, flow, &rsmq, reason).await?;
         }
     }

backend/windmill-api/openapi.yaml

@@ -1,7 +1,7 @@
 openapi: "3.0.3"
 
 info:
-  version: 1.421.0
+  version: 1.424.0
   title: Windmill API
 
   contact:
@@ -190,7 +190,7 @@ paths:
               schema:
                 type: string
 
-  /w/{workspace}/users/{username}:
+  /w/{workspace}/users/get/{username}:
     get:
       summary: get user (require admin privilege)
       operationId: getUser
@@ -283,6 +283,71 @@ paths:
             text/plain:
               schema:
                 type: string
+                
+  /users/set_password_of/{user}:
+    post:
+      summary: set password for a specific user (require super admin)
+      operationId: setPasswordForUser
+      tags:
+        - user
+      parameters:
+        - name: user
+          in: path
+          required: true
+          schema:
+            type: string
+      requestBody:
+        description: set password
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                password:
+                  type: string
+              required:
+                - password
+      responses:
+        "200":
+          description: password set
+          content:
+            text/plain:
+              schema:
+                type: string
+
+  /users/set_login_type/{user}:
+    post:
+      summary: set login type for a specific user (require super admin)
+      operationId: setLoginTypeForUser
+      tags:
+        - user
+      parameters:
+        - name: user
+          in: path
+          required: true
+          schema:
+            type: string
+      requestBody:
+        description: set login type
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                login_type:
+                  type: string
+              required:
+                - login_type
+      responses:
+        "200":
+          description: login type set
+          content:
+            text/plain:
+              schema:
+                type: string
+
 
   /users/create:
     post:
@@ -773,6 +838,78 @@ paths:
               schema:
                 type: string
 
+  /settings/critical_alerts:
+    get:
+      summary: Get all critical alerts
+      operationId: getCriticalAlerts
+      tags:
+        - setting
+      parameters:
+        - in: query
+          name: page
+          schema:
+            type: integer
+            default: 1
+            description: The page number to retrieve (minimum value is 1)
+        - in: query
+          name: page_size
+          schema:
+            type: integer
+            default: 10
+            maximum: 100
+            description: Number of alerts per page (maximum is 100)
+        - in: query
+          name: acknowledged
+          schema:
+            type: boolean
+            nullable: true
+            description: Filter by acknowledgment status; true for acknowledged, false for unacknowledged, and omit for all alerts
+      responses:
+        "200":
+          description: Successfully retrieved all critical alerts
+          content:
+            application/json:
+              schema:
+                type: array
+                items:
+                  $ref: '#/components/schemas/CriticalAlert'
+
+  /settings/critical_alerts/{id}/acknowledge:
+    post:
+      summary: Acknowledge a critical alert
+      operationId: acknowledgeCriticalAlert
+      tags:
+        - setting
+      parameters:
+        - in: path
+          name: id
+          required: true
+          schema:
+            type: integer
+          description: The ID of the critical alert to acknowledge
+      responses:
+        "200":
+          description: Successfully acknowledged the critical alert
+          content:
+            application/json:
+              schema:
+                type: string
+                example: "Critical alert acknowledged"
+
+  /settings/critical_alerts/acknowledge_all:
+    post:
+      summary: Acknowledge all unacknowledged critical alerts
+      operationId: acknowledgeAllCriticalAlerts
+      tags:
+        - setting
+      responses:
+        "200":
+          description: Successfully acknowledged all unacknowledged critical alerts.
+          content:
+            application/json:
+              schema:
+                type: string
+                example: "All unacknowledged critical alerts acknowledged"
 
   /settings/test_license_key:
     post:
@@ -1544,8 +1681,8 @@ paths:
                     type: string
                   deploy_to:
                     type: string
-                  openai_resource_path:
-                    type: string
+                  ai_resource:
+                    $ref: "#/components/schemas/AiResource"
                   code_completion_enabled:
                     type: boolean
                   error_handler:
@@ -1564,6 +1701,8 @@ paths:
                     type: string
                   default_scripts:
                     $ref: "#/components/schemas/WorkspaceDefaultScripts"
+                  mute_critical_alerts:
+                    type: boolean
                 required:
                   - code_completion_enabled
                   - automatic_billing
@@ -1829,8 +1968,8 @@ paths:
               required:
                 - code_completion_enabled
               properties:
-                openai_resource_path:
-                  type: string
+                ai_resource:
+                  $ref: "#/components/schemas/AiResource"
                 code_completion_enabled:
                   type: boolean
       responses:
@@ -1858,12 +1997,15 @@ paths:
               schema:
                 type: object
                 properties:
-                  exists_openai_resource_path:
+                  ai_provider:
+                    type: string
+                  exists_ai_resource:
                     type: boolean
                   code_completion_enabled:
                     type: boolean
                 required:
-                  - exists_openai_resource_path
+                  - ai_provider
+                  - exists_ai_resource
                   - code_completion_enabled
 
   /w/{workspace}/workspaces/edit_error_handler:
@@ -2577,6 +2719,112 @@ paths:
                 items:
                   $ref: "#/components/schemas/ContextualVariable"
 
+  /w/{workspace}/workspaces/critical_alerts:
+    get:
+      summary: Get all critical alerts for this workspace
+      operationId: workspaceGetCriticalAlerts
+      tags:
+        - setting
+      parameters:
+        - $ref: "#/components/parameters/WorkspaceId"
+        - in: query
+          name: page
+          schema:
+            type: integer
+            default: 1
+            description: The page number to retrieve (minimum value is 1)
+        - in: query
+          name: page_size
+          schema:
+            type: integer
+            default: 10
+            maximum: 100
+            description: Number of alerts per page (maximum is 100)
+        - in: query
+          name: acknowledged
+          schema:
+            type: boolean
+            nullable: true
+            description: Filter by acknowledgment status; true for acknowledged, false for unacknowledged, and omit for all alerts
+      responses:
+        "200":
+          description: Successfully retrieved all critical alerts
+          content:
+            application/json:
+              schema:
+                type: array
+                items:
+                  $ref: '#/components/schemas/CriticalAlert'
+
+  /w/{workspace}/workspaces/critical_alerts/{id}/acknowledge:
+    post:
+      summary: Acknowledge a critical alert for this workspace
+      operationId: workspaceAcknowledgeCriticalAlert
+      tags:
+        - setting
+      parameters:
+        - $ref: "#/components/parameters/WorkspaceId"
+        - in: path
+          name: id
+          required: true
+          schema:
+            type: integer
+          description: The ID of the critical alert to acknowledge
+      responses:
+        "200":
+          description: Successfully acknowledged the critical alert
+          content:
+            application/json:
+              schema:
+                type: string
+                example: "Critical alert acknowledged"
+
+  /w/{workspace}/workspaces/critical_alerts/acknowledge_all:
+    post:
+      summary: Acknowledge all unacknowledged critical alerts for this workspace
+      operationId: workspaceAcknowledgeAllCriticalAlerts
+      tags:
+        - setting
+      parameters:
+        - $ref: "#/components/parameters/WorkspaceId"
+      responses:
+        "200":
+          description: Successfully acknowledged all unacknowledged critical alerts.
+          content:
+            application/json:
+              schema:
+                type: string
+                example: "All unacknowledged critical alerts acknowledged"
+
+  /w/{workspace}/workspaces/critical_alerts/mute:
+    post:
+      summary: Mute critical alert UI for this workspace
+      operationId: workspaceMuteCriticalAlertsUI
+      tags:
+        - setting
+      parameters:
+        - $ref: "#/components/parameters/WorkspaceId"
+      requestBody:
+        description: Boolean flag to mute critical alerts.
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                mute_critical_alerts:
+                  type: boolean
+                  description: Whether critical alerts should be muted.
+                  example: true
+      responses:
+        '200':
+          description: Successfully updated mute critical alert settings.
+          content:
+            application/json:
+              schema:
+                type: string
+                example: "Updated mute critical alert UI settings for workspace: workspace_id"
+
   /oauth/login_callback/{client_name}:
     post:
       security: []
@@ -5914,6 +6162,39 @@ paths:
                 required:
                   - database_length
 
+  /w/{workspace}/jobs/completed/count_jobs:
+    get:
+      summary: count number of completed jobs with filter
+      operationId: countCompletedJobs
+      tags:
+        - job
+      parameters:
+        - $ref: "#/components/parameters/WorkspaceId"
+        - name: completed_after_s_ago
+          in: query
+          schema:
+            type: integer
+        - name: success
+          in: query
+          schema:
+            type: boolean
+        - name: tags
+          in: query
+          schema:
+            type: string
+        - name: all_workspaces
+          in: query
+          schema:
+            type: boolean
+      responses:
+        "200":
+          description: Count of completed jobs
+          content:
+            application/json:
+              schema:
+                type: integer
+
+
   /w/{workspace}/jobs/queue/list_filtered_uuids:
     get:
       summary: get the ids of all jobs matching the given filters
@@ -8116,6 +8397,23 @@ paths:
                     - id
                     - values
 
+  /workers/queue_counts:
+    get:
+      summary: get counts of jobs waiting for an executor per tag
+      operationId: getCountsOfJobsWaitingPerTag
+      tags:
+        - worker
+      responses:
+        "200":
+          description: queue counts
+          content:
+            application/json:
+              schema: 
+                type: object
+                additionalProperties:
+                  type: integer
+
+              
   /configs/list_worker_groups:
     get:
       summary: list worker groups
@@ -10040,6 +10338,16 @@ components:
   schemas:
     $ref: "../../openflow.openapi.yaml#/components/schemas"
 
+    AiResource:
+      type: object
+      properties:
+        path:
+          type: string
+        provider:
+          type: string
+      required:
+        - path
+        - provider
     Script:
       type: object
       properties:
@@ -10858,7 +11166,7 @@ components:
             - "jobs.disapproval"
             - "jobs.delete"
             - "account.delete"
-            - "openai.request"
+            - "ai.request"
             - "resources.create"
             - "resources.update"
             - "resources.delete"
@@ -12114,6 +12422,10 @@ components:
           type: object
           additionalProperties:
             type: object
+        s3_inputs:
+          type: array
+          items:
+            type: object
         execution_mode:
           type: string
           enum: [viewer, publisher, anonymous]
@@ -12721,3 +13033,27 @@ components:
           type: string
           format: date-time
 
+    CriticalAlert:
+      type: object
+      properties:
+        id:
+          type: integer
+          description: Unique identifier for the alert
+        alert_type:
+          type: string
+          description: Type of alert (e.g., critical_error)
+        message:
+          type: string
+          description: The message content of the alert
+        created_at:
+          type: string
+          format: date-time
+          description: Time when the alert was created
+        acknowledged:
+          type: boolean
+          nullable: true
+          description: Acknowledgment status of the alert, can be true, false, or null if not set
+        workspace_id:
+          type: string
+          nullable: true
+          description: Workspace id if the alert is in the scope of a workspace

backend/windmill-api/src/ai.rs

@@ -0,0 +1,507 @@
+use crate::{
+    db::{ApiAuthed, DB},
+    variables::decrypt,
+};
+use anthropic::AnthropicCache;
+use axum::{
+    body::Bytes,
+    extract::{Path, Query},
+    response::IntoResponse,
+    routing::post,
+    Extension, Router,
+};
+use lazy_static::lazy_static;
+use mistral::MistralCache;
+use openai::OpenaiCache;
+use quick_cache::sync::Cache;
+use reqwest::{Client, RequestBuilder};
+use serde::{Deserialize, Deserializer};
+use windmill_audit::audit_ee::audit_log;
+use windmill_audit::ActionKind;
+use windmill_common::error::{to_anyhow, Result};
+use windmill_common::variables::build_crypt;
+
+use windmill_common::error::Error;
+
+use serde_json::value::{RawValue, Value};
+use std::collections::HashMap;
+
+lazy_static::lazy_static! {
+    static ref HTTP_CLIENT: Client = reqwest::ClientBuilder::new()
+        .timeout(std::time::Duration::from_secs(60 * 5))
+        .user_agent("windmill/beta")
+        .build().unwrap();
+}
+
+trait AiRequest {
+    fn prepare_request(self, path: &str, body: Bytes) -> Result<RequestBuilder>;
+}
+
+mod openai {
+    use super::*;
+
+    use super::{get_variable_or_self, KeyCache};
+
+    const API_VERSION: &str = "2023-05-15";
+
+    #[derive(Deserialize, Debug)]
+    struct OpenaiResource {
+        api_key: String,
+        organization_id: Option<String>,
+    }
+
+    #[derive(Deserialize, Debug)]
+    struct OpenaiClientCredentialsOauthResource {
+        client_id: String,
+        client_secret: String,
+        token_url: String,
+        user: Option<String>,
+    }
+
+    #[derive(Deserialize, Debug)]
+    #[serde(untagged, rename_all = "snake_case")]
+    enum OpenaiConfig {
+        Resource(OpenaiResource),
+        ClientCredentialsOauthResource(OpenaiClientCredentialsOauthResource),
+    }
+
+    lazy_static::lazy_static! {
+        pub static ref OPENAI_AZURE_BASE_PATH: Option<String> = std::env::var("OPENAI_AZURE_BASE_PATH").ok();
+    }
+
+    #[derive(Deserialize, Debug)]
+    struct OpenaiCredentials {
+        access_token: String,
+    }
+
+    #[derive(Clone, Debug, Deserialize)]
+    pub struct OpenaiCache {
+        api_key: String,
+        organization_id: Option<String>,
+        azure_base_path: Option<String>,
+        user: Option<String>,
+    }
+
+    impl OpenaiCache {
+        pub fn new(
+            api_key: String,
+            organization_id: Option<String>,
+            azure_base_path: Option<String>,
+            user: Option<String>,
+        ) -> Self {
+            Self { api_key, organization_id, azure_base_path, user }
+        }
+    }
+
+    const BASE_URL: &str = "https://api.openai.com/v1";
+    impl AiRequest for OpenaiCache {
+        fn prepare_request(self, openai_path: &str, mut body: Bytes) -> Result<RequestBuilder> {
+            let OpenaiCache { api_key, azure_base_path, organization_id, user } = self;
+            if user.is_some() {
+                tracing::debug!("Adding user to request body");
+                let mut json_body: HashMap<String, Box<RawValue>> = serde_json::from_slice(&body)
+                    .map_err(|e| {
+                    Error::InternalErr(format!("Failed to parse request body: {}", e))
+                })?;
+
+                let user_json_string = serde_json::Value::String(user.unwrap()).to_string(); // makes sure to escape characters
+
+                json_body.insert(
+                    "user".to_string(),
+                    RawValue::from_string(user_json_string)
+                        .map_err(|e| Error::InternalErr(format!("Failed to parse user: {}", e)))?,
+                );
+
+                body = serde_json::to_vec(&json_body)
+                    .map_err(|e| {
+                        Error::InternalErr(format!("Failed to reserialize request body: {}", e))
+                    })?
+                    .into();
+            }
+
+            let base_url = if let Some(base_url) = azure_base_path {
+                base_url
+            } else {
+                BASE_URL.to_string()
+            };
+            let url = format!("{}/{}", base_url, openai_path);
+            let mut request = HTTP_CLIENT
+                .post(url)
+                .header("content-type", "application/json")
+                .body(body);
+
+            if base_url != BASE_URL {
+                request = request
+                    .header("api-key", api_key)
+                    .query(&[("api-version", API_VERSION)])
+            } else {
+                request = request.header("authorization", format!("Bearer {}", api_key))
+            }
+
+            if let Some(org_id) = organization_id {
+                request = request.header("OpenAI-Organization", org_id);
+            }
+
+            Ok(request)
+        }
+    }
+
+    async fn get_openai_key_using_credentials_flow(
+        mut resource: OpenaiClientCredentialsOauthResource,
+        db: &DB,
+        w_id: &str,
+    ) -> Result<String> {
+        resource.client_id = get_variable_or_self(resource.client_id, db, w_id).await?;
+        resource.client_secret = get_variable_or_self(resource.client_secret, db, w_id).await?;
+        resource.token_url = get_variable_or_self(resource.token_url, db, w_id).await?;
+        let mut params = HashMap::new();
+        params.insert("grant_type", "client_credentials");
+        let response = HTTP_CLIENT
+            .post(resource.token_url)
+            .form(&params)
+            .basic_auth(resource.client_id, Some(resource.client_secret))
+            .send()
+            .await
+            .map_err(|err| {
+                Error::InternalErr(format!(
+                    "Failed to get OpenAI credentials using credentials flow: {}",
+                    err
+                ))
+            })?;
+        let response = response.json::<OpenaiCredentials>().await.map_err(|err| {
+            Error::InternalErr(format!(
+                "Failed to parse OpenAI credentials from credentials flow: {}",
+                err
+            ))
+        })?;
+        Ok(response.access_token)
+    }
+
+    pub async fn get_cached_value(db: &DB, w_id: &str, resource: Value) -> Result<KeyCache> {
+        let config = serde_json::from_value(resource)
+            .map_err(|e| Error::InternalErr(format!("validating openai resource {e:#}")))?;
+
+        let mut user = None::<String>;
+        let mut resource = match config {
+            OpenaiConfig::Resource(resource) => {
+                tracing::debug!("Getting OpenAI key from static resource");
+                resource
+            }
+            OpenaiConfig::ClientCredentialsOauthResource(resource) => {
+                tracing::debug!("Getting OpenAI key with client credentials flow");
+                user = resource.user.clone();
+                let token = get_openai_key_using_credentials_flow(resource, db, w_id).await?;
+                OpenaiResource { api_key: token, organization_id: None }
+            }
+        };
+
+        resource.api_key = get_variable_or_self(resource.api_key, db, w_id).await?;
+
+        if let Some(organization_id) = resource.organization_id {
+            resource.organization_id = Some(get_variable_or_self(organization_id, db, w_id).await?);
+        }
+
+        if user.is_some() {
+            user = Some(get_variable_or_self(user.unwrap(), db, w_id).await?);
+        }
+
+        let azure_base_path = sqlx::query_scalar!(
+            "SELECT value
+            FROM global_settings
+            WHERE name = 'openai_azure_base_path'",
+        )
+        .fetch_optional(db)
+        .await?;
+
+        let azure_base_path = if let Some(azure_base_path) = azure_base_path {
+            Some(
+                serde_json::from_value::<String>(azure_base_path).map_err(|e| {
+                    Error::InternalErr(format!("validating openai azure base path {e:#}"))
+                })?,
+            )
+        } else {
+            OPENAI_AZURE_BASE_PATH.clone()
+        };
+
+        let workspace_cache = OpenaiCache::new(
+            resource.api_key.clone(),
+            resource.organization_id.clone(),
+            azure_base_path.clone(),
+            user.clone(),
+        );
+        Ok(KeyCache::Openai(workspace_cache))
+    }
+}
+
+mod anthropic {
+
+    use super::*;
+
+    #[derive(Clone, Deserialize, Debug)]
+    pub struct AnthropicCache {
+        #[serde(rename = "apiKey")]
+        pub api_key: String,
+    }
+
+    const API_VERSION: &str = "2023-06-01";
+
+    impl AnthropicCache {
+        pub fn new(api_key: String) -> Self {
+            Self { api_key }
+        }
+    }
+
+    const BASE_URL: &str = "https://api.anthropic.com";
+    impl AiRequest for AnthropicCache {
+        fn prepare_request(self, anthropic_path: &str, body: Bytes) -> Result<RequestBuilder> {
+            let AnthropicCache { api_key } = self;
+            let url = format!("{}/{}", BASE_URL, anthropic_path);
+            let request = HTTP_CLIENT
+                .post(url)
+                .header("x-api-key", api_key)
+                .header("anthropic-version", API_VERSION)
+                .header("content-type", "application/json")
+                .body(body);
+            Ok(request)
+        }
+    }
+
+    pub async fn get_cached_value(db: &DB, w_id: &str, resource: Value) -> Result<KeyCache> {
+        let mut resource: AnthropicCache = serde_json::from_value(resource)
+            .map_err(|e| Error::InternalErr(format!("validating anthropic resource {e:#}")))?;
+        resource.api_key = get_variable_or_self(resource.api_key, db, w_id).await?;
+        let workspace_cache = AnthropicCache::new(resource.api_key);
+        Ok(KeyCache::Anthropic(workspace_cache))
+    }
+}
+
+mod mistral {
+    use super::*;
+    #[derive(Deserialize, Clone, Debug)]
+    pub struct MistralCache {
+        #[serde(rename = "apiKey")]
+        pub api_key: String,
+    }
+
+    impl MistralCache {
+        pub fn new(api_key: String) -> Self {
+            Self { api_key }
+        }
+    }
+
+    const BASE_URL: &str = "https://api.mistral.ai";
+    impl AiRequest for MistralCache {
+        fn prepare_request(self, mistral_path: &str, body: Bytes) -> Result<RequestBuilder> {
+            let MistralCache { api_key } = self;
+
+            let url = format!("{}/{}", BASE_URL, mistral_path);
+            let request = HTTP_CLIENT
+                .post(url)
+                .header("content-type", "application/json")
+                .header("Accept", "application/json")
+                .header("authorization", format!("Bearer {}", api_key))
+                .body(body);
+            Ok(request)
+        }
+    }
+
+    pub async fn get_cached_value(db: &DB, w_id: &str, resource: Value) -> Result<KeyCache> {
+        let mut resource: MistralCache = serde_json::from_value(resource)
+            .map_err(|e| Error::InternalErr(format!("validating mistral resource {e:#}")))?;
+        resource.api_key = get_variable_or_self(resource.api_key, db, w_id).await?;
+
+        let workspace_cache = MistralCache::new(resource.api_key);
+        Ok(KeyCache::Mistral(workspace_cache))
+    }
+}
+
+#[derive(Clone, Debug)]
+pub enum KeyCache {
+    Openai(OpenaiCache),
+    Anthropic(AnthropicCache),
+    Mistral(MistralCache),
+}
+
+#[derive(Clone, Debug)]
+pub struct AiCache {
+    pub path: String,
+    pub cached_key: KeyCache,
+    pub expires_at: std::time::Instant,
+}
+
+impl AiCache {
+    pub fn new(path: String, cached_key: KeyCache) -> Self {
+        Self {
+            path,
+            cached_key,
+            expires_at: std::time::Instant::now() + std::time::Duration::from_secs(60),
+        }
+    }
+    fn is_expired(&self) -> bool {
+        self.expires_at < std::time::Instant::now()
+    }
+}
+
+lazy_static! {
+    pub static ref AI_KEY_CACHE: Cache<String, AiCache> = Cache::new(500);
+}
+
+struct Variable {
+    value: String,
+    is_secret: bool,
+}
+
+#[derive(Deserialize, Debug)]
+struct ProxyQueryParams {
+    no_cache: Option<bool>,
+}
+
+async fn get_variable_or_self(path: String, db: &DB, w_id: &str) -> Result<String> {
+    if !path.starts_with("$var:") {
+        return Ok(path);
+    }
+    let path = path.strip_prefix("$var:").unwrap().to_string();
+    let mut variable = sqlx::query_as!(
+        Variable,
+        "SELECT value, is_secret
+        FROM variable
+        WHERE path = $1 AND workspace_id = $2",
+        &path,
+        &w_id
+    )
+    .fetch_one(db)
+    .await?;
+    if variable.is_secret {
+        let mc = build_crypt(db, w_id).await?;
+        variable.value = decrypt(&mc, variable.value)?;
+    }
+    Ok(variable.value)
+}
+
+#[derive(Deserialize, Debug)]
+pub struct AiResource {
+    pub path: String,
+    #[serde(deserialize_with = "check_if_valid_ai_provider")]
+    pub provider: String,
+}
+
+fn check_if_valid_ai_provider<'de, D>(provider: D) -> std::result::Result<String, D::Error>
+where
+    D: Deserializer<'de>,
+{
+    let provider = String::deserialize(provider)?;
+    match provider.as_str() {
+        "anthropic" | "openai" | "mistral" => Ok(provider),
+        _ => Err(serde::de::Error::custom(
+            "Only the following Ai providers are supported: openai, anthropic and mistral"
+                .to_string(),
+        )),
+    }
+}
+
+pub fn workspaced_service() -> Router {
+    let router = Router::new().route("/proxy/*ai", post(proxy));
+
+    router
+}
+
+async fn proxy(
+    authed: ApiAuthed,
+    Extension(db): Extension<DB>,
+    Path((w_id, ai_path)): Path<(String, String)>,
+    Query(query_params): Query<ProxyQueryParams>,
+    body: Bytes,
+) -> impl IntoResponse {
+    let workspace_cache = AI_KEY_CACHE.get(&w_id);
+    let ai_cache = match workspace_cache {
+        Some(cache) if !cache.is_expired() && !query_params.no_cache.unwrap_or(false) => {
+            cache.cached_key
+        }
+        _ => {
+            let ai_resource = sqlx::query_scalar!(
+                "SELECT ai_resource FROM workspace_settings WHERE workspace_id = $1",
+                &w_id
+            )
+            .fetch_one(&db)
+            .await?;
+
+            if ai_resource.is_none() {
+                return Err(Error::InternalErr("AI resource not configured".to_string()));
+            }
+
+            let ai_resource = serde_json::from_value::<AiResource>(ai_resource.unwrap())
+                .map_err(|e| Error::BadRequest(e.to_string()))?;
+            let ai_resource_path = ai_resource.path;
+
+            let resource = sqlx::query_scalar!(
+                "SELECT value
+                FROM resource
+                WHERE path = $1 AND workspace_id = $2",
+                &ai_resource_path,
+                &w_id
+            )
+            .fetch_optional(&db)
+            .await?
+            .ok_or_else(|| {
+                Error::InternalErr(format!(
+                    "Could not find the {} resource at path {ai_resource_path}, update the resource path in the workspace settings", ai_resource.provider
+                ))
+            })?;
+
+            if resource.is_none() {
+                return Err(Error::InternalErr(format!(
+                    "{} resource missing value",
+                    ai_resource.provider
+                )));
+            }
+
+            let resource = resource.unwrap();
+
+            let ai_cache = match ai_resource.provider.as_str() {
+                "openai" => openai::get_cached_value(&db, &w_id, resource).await,
+                "anthropic" => anthropic::get_cached_value(&db, &w_id, resource).await,
+                "mistral" => mistral::get_cached_value(&db, &w_id, resource).await,
+                provider => {
+                    return Err(Error::BadRequest(format!("{} is not supported", provider)))
+                }
+            };
+            let ai_cache = ai_cache?;
+            AI_KEY_CACHE.insert(
+                w_id.clone(),
+                AiCache::new(ai_resource_path, ai_cache.clone()),
+            );
+            ai_cache
+        }
+    };
+    let (path, request) = match ai_cache {
+        KeyCache::Openai(cached) => ("openai_path", cached.prepare_request(&ai_path, body)),
+        KeyCache::Anthropic(cached) => ("anthropic_path", cached.prepare_request(&ai_path, body)),
+        KeyCache::Mistral(cached) => ("mistral_path", cached.prepare_request(&ai_path, body)),
+    };
+
+    let response = request?.send().await.map_err(to_anyhow)?;
+
+    let mut tx = db.begin().await?;
+
+    audit_log(
+        &mut *tx,
+        &authed,
+        "ai.request",
+        ActionKind::Execute,
+        &w_id,
+        Some(&authed.email),
+        Some([(path, &format!("{:?}", ai_path)[..])].into()),
+    )
+    .await?;
+    tx.commit().await?;
+
+    if response.error_for_status_ref().is_err() {
+        let err_msg = response.text().await.unwrap_or("".to_string());
+        return Err(Error::AiError(err_msg));
+    }
+
+    let status_code = response.status();
+    let headers = response.headers().clone();
+    let stream = response.bytes_stream();
+    Ok((status_code, headers, axum::body::Body::from_stream(stream)))
+}

backend/windmill-api/src/apps.rs

@@ -7,6 +7,12 @@ use std::collections::HashMap;
  * Please see the included NOTICE for copyright information and
  * LICENSE-AGPL for a copy of the license.
  */
+
+#[cfg(feature = "parquet")]
+use crate::{job_helpers_ee::{
+    get_random_file_name, get_s3_resource, get_workspace_s3_resource, upload_file_internal,
+    UploadFileResponse,
+}, users::fetch_api_authed_from_permissioned_as};
 use crate::{
     db::{ApiAuthed, DB},
     resources::get_resource_value_interpolated_internal,
@@ -23,7 +29,13 @@ use axum::{
     Router,
 };
 use hyper::StatusCode;
+#[cfg(feature = "parquet")]
+use itertools::Itertools;
 use magic_crypt::MagicCryptTrait;
+#[cfg(feature = "parquet")]
+use object_store::{Attribute, Attributes};
+#[cfg(feature = "parquet")]
+use regex::Regex;
 use serde::{Deserialize, Serialize};
 use serde_json::{json, value::RawValue};
 use sha2::{Digest, Sha256};
@@ -32,6 +44,8 @@ use sqlx::{types::Uuid, FromRow};
 use std::str;
 use windmill_audit::audit_ee::audit_log;
 use windmill_audit::ActionKind;
+#[cfg(feature = "parquet")]
+use windmill_common::s3_helpers::build_object_store_client;
 use windmill_common::{
     apps::ListAppQuery,
     db::UserDB,
@@ -69,6 +83,7 @@ pub fn workspaced_service() -> Router {
 pub fn unauthed_service() -> Router {
     Router::new()
         .route("/execute_component/*path", post(execute_component))
+        .route("/upload_s3_file/*path", post(upload_s3_file_from_app))
         .route("/public_app/:secret", get(get_public_app_by_secret))
         .route("/public_resource/*path", get(get_public_resource))
 }
@@ -179,6 +194,14 @@ pub struct PolicyTriggerableInputs {
     allow_user_resources: AllowUserResources,
 }
 
+#[derive(Serialize, Deserialize, Debug, Clone)]
+pub struct S3Input {
+    allowed_resources: Vec<String>,
+    allow_user_resources: bool,
+    allow_workspace_resource: bool,
+    file_key_regex: String,
+}
+
 #[derive(Serialize, Deserialize, Debug, Clone)]
 pub struct Policy {
     pub on_behalf_of: Option<String>,
@@ -192,6 +215,7 @@ pub struct Policy {
     #[serde(skip_serializing_if = "Option::is_none")]
     pub triggerables_v2: Option<HashMap<String, PolicyTriggerableInputs>>,
     pub execution_mode: ExecutionMode,
+    pub s3_inputs: Option<Vec<S3Input>>,
 }
 
 #[derive(Deserialize)]
@@ -432,9 +456,7 @@ async fn get_latest_version(
     authed: ApiAuthed,
     Extension(user_db): Extension<UserDB>,
     Path((w_id, path)): Path<(String, StripPath)>,
-
 ) -> JsonResult<Option<AppHistory>> {
-
     let mut tx = user_db.begin(&authed).await?;
     let row = sqlx::query!(
         "SELECT a.id as app_id, av.id as version_id, dm.deployment_msg as deployment_msg
@@ -457,7 +479,6 @@ async fn get_latest_version(
     } else {
         return Ok(Json(None));
     }
-
 }
 
 async fn update_app_history(
@@ -1067,6 +1088,49 @@ fn digest(code: &str) -> String {
     format!("rawscript/{:x}", result)
 }
 
+async fn get_on_behalf_details_from_policy_and_authed(
+    policy: &Policy,
+    opt_authed: &Option<ApiAuthed>,
+) -> Result<(String, String, String)> {
+    let (username, permissioned_as, email) = match policy.execution_mode {
+        ExecutionMode::Anonymous => {
+            let username = opt_authed
+                .as_ref()
+                .map(|a| a.username.clone())
+                .unwrap_or_else(|| "anonymous".to_string());
+            let (permissioned_as, email) = get_on_behalf_of(&policy)?;
+            (username, permissioned_as, email)
+        }
+        ExecutionMode::Publisher => {
+            let username = opt_authed
+                .as_ref()
+                .map(|a| a.username.clone())
+                .ok_or_else(|| {
+                    Error::BadRequest(
+                        "publisher execution mode requires authentication".to_string(),
+                    )
+                })?;
+            let (permissioned_as, email) = get_on_behalf_of(&policy)?;
+            (username, permissioned_as, email)
+        }
+        ExecutionMode::Viewer => {
+            let (username, email) = opt_authed
+                .as_ref()
+                .map(|a| (a.username.clone(), a.email.clone()))
+                .ok_or_else(|| {
+                    Error::BadRequest("Required to be authed in viewer mode".to_string())
+                })?;
+            (
+                username.clone(),
+                username_to_permissioned_as(&username),
+                email,
+            )
+        }
+    };
+
+    Ok((username, permissioned_as, email))
+}
+
 async fn execute_component(
     OptAuthed(opt_authed): OptAuthed,
     Extension(db): Extension<DB>,
@@ -1129,6 +1193,7 @@ async fn execute_component(
                 triggerables_v2: Some(hm),
                 on_behalf_of: None,
                 on_behalf_of_email: None,
+                s3_inputs: None,
             }
         }
         _ => {
@@ -1146,41 +1211,8 @@ async fn execute_component(
         }
     };
 
-    let (username, permissioned_as, email) = match policy.execution_mode {
-        ExecutionMode::Anonymous => {
-            let username = opt_authed
-                .as_ref()
-                .map(|a| a.username.clone())
-                .unwrap_or_else(|| "anonymous".to_string());
-            let (permissioned_as, email) = get_on_behalf_of(&policy)?;
-            (username, permissioned_as, email)
-        }
-        ExecutionMode::Publisher => {
-            let username = opt_authed
-                .as_ref()
-                .map(|a| a.username.clone())
-                .ok_or_else(|| {
-                    Error::BadRequest(
-                        "publisher execution mode requires authentication".to_string(),
-                    )
-                })?;
-            let (permissioned_as, email) = get_on_behalf_of(&policy)?;
-            (username, permissioned_as, email)
-        }
-        ExecutionMode::Viewer => {
-            let (username, email) = opt_authed
-                .as_ref()
-                .map(|a| (a.username.clone(), a.email.clone()))
-                .ok_or_else(|| {
-                    Error::BadRequest("Required to be authed in viewer mode".to_string())
-                })?;
-            (
-                username.clone(),
-                username_to_permissioned_as(&username),
-                email,
-            )
-        }
-    };
+    let (username, permissioned_as, email) =
+        get_on_behalf_details_from_policy_and_authed(&policy, &opt_authed).await?;
 
     let (job_payload, (args, job_id), tag) = match payload {
         ExecuteApp { args, component, raw_code: Some(raw_code), path: None, .. } => {
@@ -1249,6 +1281,248 @@ async fn execute_component(
     Ok(uuid.to_string())
 }
 
+#[cfg(not(feature = "parquet"))]
+async fn upload_s3_file_from_app() -> Result<()> {
+    return Err(Error::BadRequest(
+        "This endpoint requires the parquet feature to be enabled".to_string(),
+    ));
+}
+
+
+#[cfg(feature = "parquet")]
+#[derive(Debug, Deserialize, Clone)]
+struct UploadFileToS3Query {
+    file_key: Option<String>,
+    file_extension: Option<String>,
+    s3_resource_path: Option<String>,
+    content_type: Option<String>,
+    content_disposition: Option<String>,
+    force_viewer_file_key_regex: Option<String>,
+    force_viewer_allow_user_resources: Option<bool>,
+    force_viewer_allow_workspace_resource: Option<bool>,
+    force_viewer_allowed_resources: Option<String>,
+}
+
+#[cfg(feature = "parquet")]
+async fn upload_s3_file_from_app(
+    OptAuthed(opt_authed): OptAuthed,
+    Extension(db): Extension<DB>,
+    Path((w_id, path)): Path<(String, StripPath)>,
+    Query(query): Query<UploadFileToS3Query>,
+    request: axum::extract::Request,
+) -> JsonResult<UploadFileResponse> {
+    let policy = if let Some(file_key_regex) = query.force_viewer_file_key_regex {
+        Some(Policy {
+            execution_mode: ExecutionMode::Viewer,
+            triggerables: None,
+            triggerables_v2: None,
+            on_behalf_of: None,
+            on_behalf_of_email: None,
+            s3_inputs: Some(vec![S3Input {
+                file_key_regex: file_key_regex,
+                allow_user_resources: query.force_viewer_allow_user_resources.unwrap_or(false),
+                allow_workspace_resource: query
+                    .force_viewer_allow_workspace_resource
+                    .unwrap_or(false),
+                allowed_resources: query
+                    .force_viewer_allowed_resources
+                    .map(|s| s.split(',').map(|s| s.to_string()).collect())
+                    .unwrap_or_default(),
+            }]),
+        })
+    } else {
+        let policy_o = sqlx::query_scalar!(
+            "SELECT policy from app WHERE path = $1 AND workspace_id = $2",
+            &path.0,
+            &w_id
+        )
+        .fetch_optional(&db)
+        .await?;
+
+        policy_o
+            .map(|p| serde_json::from_value::<Policy>(p).map_err(to_anyhow))
+            .transpose()?
+    };
+
+    let user_db = UserDB::new(db.clone());
+
+    let (s3_resource_opt, file_key) = if policy.as_ref().is_some_and(|p| p.s3_inputs.is_some()) {
+        let policy = policy.unwrap();
+        let s3_inputs = policy.s3_inputs.as_ref().unwrap();
+
+        let (username, permissioned_as, email) =
+            get_on_behalf_details_from_policy_and_authed(&policy, &opt_authed).await?;
+
+        let on_behalf_authed =
+            fetch_api_authed_from_permissioned_as(permissioned_as, email, &w_id, &db, username)
+                .await?;
+
+        if let Some(file_key) = query.file_key {
+            // file key is provided => requires workspace, user or list policy and must match the regex
+            let matching_s3_inputs = if let Some(ref s3_resource_path) = query.s3_resource_path {
+                s3_inputs
+                    .iter()
+                    .filter(|s3_input| {
+                        s3_input.allowed_resources.contains(s3_resource_path)
+                            || s3_input.allow_user_resources
+                    })
+                    .sorted_by_key(|i| i.allow_user_resources) // consider user resources last
+                    .collect::<Vec<_>>()
+            } else {
+                s3_inputs
+                    .iter()
+                    .filter(|s3_input| s3_input.allow_workspace_resource)
+                    .collect::<Vec<_>>()
+            };
+
+            let matched_input = matching_s3_inputs.iter().find(|s3_input| {
+                match Regex::new(&s3_input.file_key_regex) {
+                    Ok(re) => re.is_match(&file_key),
+                    Err(e) => {
+                        tracing::error!("Error compiling regex: {}", e);
+                        false
+                    }
+                }
+            });
+
+            if let Some(matched_input) = matched_input {
+                if let Some(ref s3_resource_path) = query.s3_resource_path {
+                    if matched_input.allow_user_resources {
+                        if let Some(authed) = opt_authed {
+                            (
+                                Some(
+                                    get_s3_resource(
+                                        &authed,
+                                        &db,
+                                        Some(user_db),
+                                        "",
+                                        &w_id,
+                                        s3_resource_path,
+                                        None,
+                                        None,
+                                    )
+                                    .await?,
+                                ),
+                                file_key,
+                            )
+                        } else {
+                            return Err(Error::BadRequest(
+                                "User resources are not allowed without being logged in"
+                                    .to_string(),
+                            ));
+                        }
+                    } else {
+                        (
+                            Some(
+                                get_s3_resource(
+                                    &on_behalf_authed,
+                                    &db,
+                                    Some(user_db),
+                                    "",
+                                    &w_id,
+                                    s3_resource_path,
+                                    None,
+                                    None,
+                                )
+                                .await?,
+                            ),
+                            file_key,
+                        )
+                    }
+                } else {
+                    let (_, s3_resource_opt) =
+                        get_workspace_s3_resource(&on_behalf_authed, &db, None, "", &w_id, None)
+                            .await?;
+                    (s3_resource_opt, file_key)
+                }
+            } else {
+                return Err(Error::BadRequest(
+                    "No matching s3 resource found for the given file key".to_string(),
+                ));
+            }
+        } else {
+            // no file key => requires unnamed upload policy => allow workspace resource and file_key_regex is empty
+            let has_unnamed_policy = s3_inputs.iter().any(|s3_input| {
+                s3_input.allow_workspace_resource && s3_input.file_key_regex.is_empty()
+            });
+
+            if !has_unnamed_policy {
+                return Err(Error::BadRequest(
+                    "no policy found for unnamed s3 file uplooad".to_string(),
+                ));
+            }
+
+            // for now, we place all files into `windmill_uploads` folder with a random name
+            // TODO: make the folder configurable via the workspace settings
+            let file_key = get_random_file_name(query.file_extension);
+
+            let (_, s3_resource_opt) =
+                get_workspace_s3_resource(&on_behalf_authed, &db, None, "", &w_id, None).await?;
+
+            (s3_resource_opt, file_key)
+        }
+    } else {
+        // backward compatibility (no policy)
+        // if no policy but logged in, use the user's auth to get the s3 resource
+        if let Some(authed) = opt_authed {
+            let file_key = query
+                .file_key
+                .unwrap_or_else(|| get_random_file_name(query.file_extension));
+
+            if let Some(ref s3_resource_path) = query.s3_resource_path {
+                (
+                    Some(
+                        get_s3_resource(
+                            &authed,
+                            &db,
+                            Some(user_db),
+                            "",
+                            &w_id,
+                            s3_resource_path,
+                            None,
+                            None,
+                        )
+                        .await?,
+                    ),
+                    file_key,
+                )
+            } else {
+                let (_, s3_resource) =
+                    get_workspace_s3_resource(&authed, &db, None, "", &w_id, None).await?;
+
+                (s3_resource, file_key)
+            }
+        } else {
+            return Err(Error::BadRequest("Missing s3 policy".to_string()));
+        }
+    };
+
+    let s3_resource = s3_resource_opt.ok_or(Error::InternalErr(
+        "No files storage resource defined at the workspace level".to_string(),
+    ))?;
+    let s3_client = build_object_store_client(&s3_resource).await?;
+
+    let options = Attributes::from_iter(vec![
+        (
+            Attribute::ContentType,
+            query.content_type.unwrap_or_else(|| {
+                mime_guess::from_path(&file_key)
+                    .first_or_octet_stream()
+                    .to_string()
+            }),
+        ),
+        (
+            Attribute::ContentDisposition,
+            query.content_disposition.unwrap_or("inline".to_string()),
+        ),
+    ])
+    .into();
+
+    upload_file_internal(s3_client, &file_key, request, options).await?;
+
+    return Ok(Json(UploadFileResponse { file_key }));
+}
+
 fn get_on_behalf_of(policy: &Policy) -> Result<(String, String)> {
     let permissioned_as = policy
         .on_behalf_of

backend/windmill-api/src/job_helpers_ee.rs

@@ -1,9 +1,21 @@
 use axum::Router;
+use serde::Serialize;
+use uuid::Uuid;
+use windmill_common::s3_helpers::StorageResourceType;
 
 #[cfg(feature = "parquet")]
 use crate::db::{ApiAuthed, DB};
 #[cfg(feature = "parquet")]
+use object_store::{ObjectStore, PutMultipartOpts};
+#[cfg(feature = "parquet")]
+use std::sync::Arc;
+use windmill_common::error;
+#[cfg(feature = "parquet")]
 use windmill_common::{db::UserDB, s3_helpers::ObjectStoreResource};
+#[derive(Serialize)]
+pub struct UploadFileResponse {
+    pub file_key: String,
+}
 
 pub fn workspaced_service() -> Router {
     Router::new()
@@ -21,3 +33,29 @@ pub async fn get_workspace_s3_resource<'c>(
     // implementation is not open source
     Ok((None, None))
 }
+
+pub fn get_random_file_name(_file_extension: Option<String>) -> String {
+    todo!()
+}
+
+pub async fn get_s3_resource<'c>(
+    _authed: &ApiAuthed,
+    _db: &DB,
+    _user_db: Option<UserDB>,
+    _token: &str,
+    _w_id: &str,
+    _resource_path: &str,
+    _resource_type: Option<StorageResourceType>,
+    _job_id: Option<Uuid>,
+) -> error::Result<ObjectStoreResource> {
+    todo!()
+}
+
+pub async fn upload_file_internal(
+    _s3_client: Arc<dyn ObjectStore>,
+    _file_key: &str,
+    _request: axum::extract::Request,
+    _options: PutMultipartOpts,
+) -> error::Result<()> {
+    todo!()
+}

backend/windmill-api/src/jobs.rs

@@ -12,6 +12,7 @@ use quick_cache::sync::Cache;
 use serde_json::value::RawValue;
 use sqlx::Pool;
 use std::collections::HashMap;
+use std::ops::{Deref, DerefMut};
 #[cfg(feature = "prometheus")]
 use std::sync::atomic::Ordering;
 use tokio::io::AsyncReadExt;
@@ -65,7 +66,7 @@ use windmill_common::{
     db::UserDB,
     error::{self, to_anyhow, Error},
     flow_status::{Approval, FlowStatus, FlowStatusModule},
-    flows::FlowValue,
+    flows::{add_virtual_items_if_necessary, FlowValue},
     jobs::{script_path_to_payload, CompletedJob, JobKind, JobPayload, QueuedJob, RawCode},
     oauth2::HmacSha256,
     scripts::{ScriptHash, ScriptLang},
@@ -205,6 +206,7 @@ pub fn workspaced_service() -> Router {
         .route("/queue/list_filtered_uuids", get(list_filtered_uuids))
         .route("/queue/cancel_selection", post(cancel_selection))
         .route("/completed/count", get(count_completed_jobs))
+        .route("/completed/count_jobs", get(count_completed_jobs_detail))
         .route(
             "/completed/list",
             get(list_completed_jobs).layer(cors.clone()),
@@ -600,7 +602,7 @@ async fn get_flow_job_debug_info(
     Extension(db): Extension<DB>,
     Path((w_id, id)): Path<(String, Uuid)>,
 ) -> error::Result<Response> {
-    let job = get_queued_job(&id, w_id.as_str(), &db).await?;
+    let job = get_queued_job_ex(&db, &w_id, id, false, None).await?;
     if let Some(job) = job {
         let is_flow = job.is_flow();
         if job.is_flow_step || !is_flow {
@@ -734,6 +736,67 @@ fn generate_get_job_query(no_logs: bool, table: &str) -> String {
     {join}
     WHERE id = $1 AND {table}.workspace_id = $2");
 }
+pub async fn get_queued_job_ex(
+    db: &DB,
+    workspace_id: &str,
+    job_id: Uuid,
+    no_logs: bool,
+    // first optional is if authed need to be checked, second is the opt_authed itself
+    opt_authed: Option<&Option<ApiAuthed>>,
+) -> error::Result<Option<JobExtended<QueuedJob>>> {
+    let query = if no_logs { &*GET_QUEUED_JOB_QUERY_NO_LOGS } else { &*GET_QUEUED_JOB_QUERY };
+    let job = sqlx::query_as::<_, JobExtended<QueuedJob>>(query)
+        .bind(job_id)
+        .bind(workspace_id)
+        .fetch_optional(db)
+        .await?;
+
+    if let Some(job) = job.as_ref() {
+        if opt_authed.is_some_and(|x| x.is_none()) && job.created_by != "anonymous" {
+            return Err(Error::BadRequest(
+                "As a non logged in user, you can only see jobs ran by anonymous users".to_string(),
+            ));
+        }
+    }
+
+    Ok(job)
+}
+pub async fn get_completed_job_ex(
+    db: &DB,
+    workspace_id: &str,
+    job_id: Uuid,
+    no_logs: bool,
+    // first optional is if authed need to be checked, second is the opt_authed itself
+    opt_authed: Option<&Option<ApiAuthed>>,
+) -> error::Result<Option<JobExtended<CompletedJob>>> {
+    let query = if no_logs { &*GET_COMPLETED_JOB_QUERY_NO_LOGS } else { &*GET_COMPLETED_JOB_QUERY };
+    let cjob = sqlx::query_as::<_, JobExtended<CompletedJob>>(query)
+        .bind(job_id)
+        .bind(workspace_id)
+        .fetch_optional(db)
+        .await?;
+
+    if let Some(job) = cjob.as_ref() {
+        if opt_authed.is_some_and(|x| x.is_none()) && job.created_by != "anonymous" {
+            return Err(Error::BadRequest(
+                "As a non logged in user, you can only see jobs ran by anonymous users".to_string(),
+            ));
+        }
+    }
+
+    if let Some(mut cjob) = cjob {
+        let CompletedJobWithFormattedResult { mut cj, result } = format_completed_job_result(cjob.inner);
+        cj.result = match result {
+            Some(FormattedResult::RawValue(rv)) => rv,
+            Some(FormattedResult::Vec(v)) => Some(to_raw_value(&v)),
+            None => None,
+        }.map(sqlx::types::Json);
+        cjob.inner = cj;
+        return Ok(Some(cjob));
+    }
+
+    Ok(cjob)
+}
 pub async fn get_job_internal(
     db: &DB,
     workspace_id: &str,
@@ -742,48 +805,18 @@ pub async fn get_job_internal(
     // first optional is if authed need to be checked, second is the opt_authed itself
     opt_authed: Option<&Option<ApiAuthed>>,
 ) -> error::Result<Job> {
-    let cjob_maybe = sqlx::query_as::<_, CompletedJob>(if no_logs {
-        &*GET_COMPLETED_JOB_QUERY_NO_LOGS
-    } else {
-        &*GET_COMPLETED_JOB_QUERY
-    })
-    .bind(job_id)
-    .bind(workspace_id)
-    .fetch_optional(db)
-    .await?
-    .map(Job::CompletedJob);
-
-    if let Some(cjob) = cjob_maybe {
-        Ok(match cjob {
-            Job::CompletedJob(cjob) => {
-                if opt_authed.is_some_and(|x| x.is_none()) && cjob.created_by != "anonymous" {
-                    return Err(Error::BadRequest(
-                        "As a non logged in user, you can only see jobs ran by anonymous users"
-                            .to_string(),
-                    ));
-                }
-                Job::CompletedJobWithFormattedResult(format_completed_job_result(cjob))
-            }
-            cjob => cjob,
-        })
-    } else {
-        let job_o = sqlx::query_as::<_, QueuedJob>(if no_logs {
-            &*GET_QUEUED_JOB_QUERY_NO_LOGS
-        } else {
-            &*GET_QUEUED_JOB_QUERY
-        })
-        .bind(job_id)
-        .bind(workspace_id)
-        .fetch_optional(db)
+    let cjob = get_completed_job_ex(db, workspace_id, job_id, no_logs, opt_authed.clone())
         .await?
-        .map(Job::QueuedJob);
-        let job: Job = not_found_if_none(job_o, "Job", job_id.to_string())?;
-        if opt_authed.is_some_and(|x| x.is_none()) && job.created_by() != "anonymous" {
-            return Err(Error::BadRequest(
-                "As a non logged in user, you can only see jobs ran by anonymous users".to_string(),
-            ));
+        .map(Job::CompletedJob);
+
+    match cjob {
+        Some(cjob) => Ok(cjob),
+        None => {
+            let job_maybe = get_queued_job_ex(db, workspace_id, job_id, no_logs, opt_authed)
+                .await?
+                .map(Job::QueuedJob);
+            not_found_if_none(job_maybe, "Job", job_id.to_string())
         }
-        Ok(job)
     }
 }
 
@@ -1560,6 +1593,50 @@ async fn count_queue_jobs(
     ))
 }
 
+#[derive(Deserialize)]
+pub struct CountCompletedJobsQuery {
+    completed_after_s_ago: Option<i64>,
+    success: Option<bool>,
+    tags: Option<String>,
+    all_workspaces: Option<bool>,
+}
+
+async fn count_completed_jobs_detail(
+    Extension(db): Extension<DB>,
+    Path(w_id): Path<String>,
+    Query(query): Query<CountCompletedJobsQuery>,
+) -> error::JsonResult<i64> {
+    let mut sqlb = SqlBuilder::select_from("completed_job");
+    sqlb
+        .field("COUNT(*) as count");
+
+    if !query.all_workspaces.unwrap_or(false) {
+        sqlb.and_where_eq("workspace_id", "?".bind(&w_id));
+    }
+
+    if let Some(after_s_ago) = query.completed_after_s_ago {
+        let after = Utc::now() - chrono::Duration::seconds(after_s_ago);
+        sqlb.and_where_gt("started_at + duration_ms / 1000 * interval '1 second'", "?".bind(&after.to_rfc3339()));
+    }
+
+    if let Some(success) = query.success {
+        sqlb.and_where_eq("success", "?".bind(&success));
+    }
+
+    if let Some(tags) = query.tags {
+        sqlb.and_where_in("tag", &tags.split(",").map(|t| format!("'{}'", t)).collect::<Vec<_>>());
+    }
+
+    let sql = sqlb.sql()?;
+    let stats = sqlx::query_scalar::<_, i64>(&sql)
+        .fetch_one(&db)
+        .await?;
+
+    Ok(Json(stats))
+}
+
+
+
 async fn count_completed_jobs(
     Extension(db): Extension<DB>,
     Path(w_id): Path<String>,
@@ -1735,7 +1812,6 @@ async fn resume_suspended_job_internal(
     let trigger_email = match &parent_flow {
         Job::CompletedJob(job) => &job.email,
         Job::QueuedJob(job) => &job.email,
-        Job::CompletedJobWithFormattedResult(job) => &job.cj.email,
     };
     conditionally_require_authed_user(authed.clone(), flow_status, trigger_email)?;
 
@@ -2010,7 +2086,6 @@ pub async fn get_suspended_job_flow(
     let trigger_email = match &flow {
         Job::CompletedJob(job) => &job.email,
         Job::QueuedJob(job) => &job.email,
-        Job::CompletedJobWithFormattedResult(job) => &job.cj.email,
     };
     conditionally_require_authed_user(authed.clone(), flow_status.clone(), trigger_email)?;
 
@@ -2225,13 +2300,45 @@ pub async fn get_resume_urls(
     Ok(Json(res))
 }
 
+#[derive(sqlx::FromRow, Debug, Serialize)]
+pub struct JobExtended<T> {
+    #[sqlx(flatten)]
+    #[serde(flatten)]
+    inner: T,
+
+    #[sqlx(skip)]
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub self_wait_time_ms: Option<i64>,
+    #[sqlx(skip)]
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub aggregate_wait_time_ms: Option<i64>,
+}
+
+impl<T> JobExtended<T> {
+    pub fn new(self_wait_time_ms: Option<i64>, aggregate_wait_time_ms: Option<i64>, inner: T) -> Self {
+        Self { inner, self_wait_time_ms, aggregate_wait_time_ms }
+    }
+}
+
+impl<T> Deref for JobExtended<T> {
+    type Target = T;
+
+    fn deref(&self) -> &Self::Target {
+        &self.inner
+    }
+}
+
+impl<T> DerefMut for JobExtended<T> {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        &mut self.inner
+    }
+}
+
 #[derive(Serialize, Debug)]
 #[serde(tag = "type")]
 pub enum Job {
-    QueuedJob(QueuedJob),
-    CompletedJob(CompletedJob),
-    #[serde(rename = "CompletedJob")]
-    CompletedJobWithFormattedResult(CompletedJobWithFormattedResult),
+    QueuedJob(JobExtended<QueuedJob>),
+    CompletedJob(JobExtended<CompletedJob>),
 }
 
 impl Job {
@@ -2239,27 +2346,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => &job.created_by,
             Job::CompletedJob(job) => &job.created_by,
-            Job::CompletedJobWithFormattedResult(job) => &job.cj.created_by,
-        }
-    }
-    pub fn raw_flow(&self) -> Option<FlowValue> {
-        match self {
-            Job::QueuedJob(job) => job
-                .raw_flow
-                .as_ref()
-                .map(|rf| serde_json::from_str(rf.0.get()).ok())
-                .flatten(),
-            Job::CompletedJob(job) => job
-                .raw_flow
-                .as_ref()
-                .map(|rf| serde_json::from_str(rf.0.get()).ok())
-                .flatten(),
-            Job::CompletedJobWithFormattedResult(job) => job
-                .cj
-                .raw_flow
-                .as_ref()
-                .map(|rf| serde_json::from_str(rf.0.get()).ok())
-                .flatten(),
         }
     }
 
@@ -2279,13 +2365,6 @@ impl Job {
                     job.logs = Some(logs.to_string());
                 }
             }
-            Job::CompletedJobWithFormattedResult(job) => {
-                if let Some(ref mut l) = job.cj.logs {
-                    l.push_str(logs);
-                } else {
-                    job.cj.logs = Some(logs.to_string());
-                }
-            }
         }
     }
 
@@ -2293,7 +2372,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => job.logs.as_ref().map(|l| l.len()),
             Job::CompletedJob(job) => job.logs.as_ref().map(|l| l.len()),
-            Job::CompletedJobWithFormattedResult(job) => job.cj.logs.as_ref().map(|l| l.len()),
         }
     }
 
@@ -2301,7 +2379,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => job.logs.clone(),
             Job::CompletedJob(job) => job.logs.clone(),
-            Job::CompletedJobWithFormattedResult(job) => job.cj.logs.clone(),
         }
     }
     pub fn flow_status(&self) -> Option<FlowStatus> {
@@ -2316,19 +2393,12 @@ impl Job {
                 .as_ref()
                 .map(|rf| serde_json::from_str(rf.0.get()).ok())
                 .flatten(),
-            Job::CompletedJobWithFormattedResult(job) => job
-                .cj
-                .flow_status
-                .as_ref()
-                .map(|rf| serde_json::from_str(rf.0.get()).ok())
-                .flatten(),
         }
     }
     pub fn is_flow_step(&self) -> bool {
         match self {
             Job::QueuedJob(job) => job.is_flow_step,
             Job::CompletedJob(job) => job.is_flow_step,
-            Job::CompletedJobWithFormattedResult(job) => job.cj.is_flow_step,
         }
     }
 
@@ -2343,7 +2413,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => &job.job_kind,
             Job::CompletedJob(job) => &job.job_kind,
-            Job::CompletedJobWithFormattedResult(job) => &job.cj.job_kind,
         }
     }
 
@@ -2351,7 +2420,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => job.id,
             Job::CompletedJob(job) => job.id,
-            Job::CompletedJobWithFormattedResult(job) => job.cj.id,
         }
     }
 
@@ -2359,7 +2427,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => &job.workspace_id,
             Job::CompletedJob(job) => &job.workspace_id,
-            Job::CompletedJobWithFormattedResult(job) => &job.cj.workspace_id,
         }
     }
 
@@ -2367,7 +2434,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => job.script_path.as_ref(),
             Job::CompletedJob(job) => job.script_path.as_ref(),
-            Job::CompletedJobWithFormattedResult(job) => job.cj.script_path.as_ref(),
         }
         .map(String::as_str)
         .unwrap_or("tmp/main")
@@ -2377,7 +2443,6 @@ impl Job {
         match self {
             Job::QueuedJob(job) => job.args.as_ref(),
             Job::CompletedJob(job) => job.args.as_ref(),
-            Job::CompletedJobWithFormattedResult(job) => job.cj.args.as_ref(),
         }
     }
 
@@ -2426,10 +2491,6 @@ impl Job {
                 job.self_wait_time_ms = self_wait_time;
                 job.aggregate_wait_time_ms = aggregate_wait_time;
             }
-            Job::CompletedJobWithFormattedResult(job) => {
-                job.cj.self_wait_time_ms = self_wait_time;
-                job.cj.aggregate_wait_time_ms = aggregate_wait_time;
-            }
         }
         Ok(())
     }
@@ -2557,86 +2618,82 @@ impl UnifiedJob {
 impl<'a> From<UnifiedJob> for Job {
     fn from(uj: UnifiedJob) -> Self {
         match uj.typ.as_ref() {
-            "CompletedJob" => Job::CompletedJob(CompletedJob {
-                workspace_id: uj.workspace_id,
-                id: uj.id,
-                parent_job: uj.parent_job,
-                created_by: uj.created_by,
-                created_at: uj.created_at,
-                started_at: uj.started_at.unwrap_or(uj.created_at),
-                duration_ms: uj.duration_ms.unwrap(),
-                success: uj.success.unwrap(),
-                script_hash: uj.script_hash,
-                script_path: uj.script_path,
-                args: None,
-                result: None,
-                logs: None,
-                flow_status: None,
-                deleted: uj.deleted,
-                canceled: uj.canceled,
-                canceled_by: uj.canceled_by,
-                raw_code: None,
-                canceled_reason: None,
-                job_kind: uj.job_kind,
-                schedule_path: uj.schedule_path,
-                permissioned_as: uj.permissioned_as,
-                raw_flow: None,
-                is_flow_step: uj.is_flow_step,
-                language: uj.language,
-                is_skipped: uj.is_skipped,
-                email: uj.email,
-                visible_to_owner: uj.visible_to_owner,
-                mem_peak: uj.mem_peak,
-                tag: uj.tag,
-                priority: uj.priority,
-                labels: uj.labels,
-                self_wait_time_ms: uj.self_wait_time_ms,
-                aggregate_wait_time_ms: uj.aggregate_wait_time_ms,
-            }),
-            "QueuedJob" => Job::QueuedJob(QueuedJob {
-                workspace_id: uj.workspace_id,
-                id: uj.id,
-                parent_job: uj.parent_job,
-                created_by: uj.created_by,
-                created_at: uj.created_at,
-                started_at: uj.started_at,
-                script_hash: uj.script_hash,
-                script_path: uj.script_path,
-                args: None,
-                running: uj.running.unwrap(),
-                scheduled_for: uj.scheduled_for.unwrap(),
-                logs: None,
-                flow_status: None,
-                raw_code: None,
-                raw_lock: None,
-                canceled: uj.canceled,
-                canceled_by: uj.canceled_by,
-                canceled_reason: None,
-                last_ping: None,
-                job_kind: uj.job_kind,
-                schedule_path: uj.schedule_path,
-                permissioned_as: uj.permissioned_as,
-                raw_flow: None,
-                is_flow_step: uj.is_flow_step,
-                language: uj.language,
-                same_worker: false,
-                pre_run_error: None,
-                email: uj.email,
-                visible_to_owner: uj.visible_to_owner,
-                suspend: uj.suspend,
-                mem_peak: uj.mem_peak,
-                root_job: None,
-                leaf_jobs: None,
-                tag: uj.tag,
-                concurrent_limit: uj.concurrent_limit,
-                concurrency_time_window_s: uj.concurrency_time_window_s,
-                timeout: None,
-                flow_step_id: None,
-                cache_ttl: None,
-                priority: uj.priority,
-                self_wait_time_ms: uj.self_wait_time_ms,
-                aggregate_wait_time_ms: uj.aggregate_wait_time_ms,
-            }),
+            "CompletedJob" => Job::CompletedJob(JobExtended::new(uj.self_wait_time_ms, uj.aggregate_wait_time_ms, CompletedJob {
+                    workspace_id: uj.workspace_id,
+                    id: uj.id,
+                    parent_job: uj.parent_job,
+                    created_by: uj.created_by,
+                    created_at: uj.created_at,
+                    started_at: uj.started_at.unwrap_or(uj.created_at),
+                    duration_ms: uj.duration_ms.unwrap(),
+                    success: uj.success.unwrap(),
+                    script_hash: uj.script_hash,
+                    script_path: uj.script_path,
+                    args: None,
+                    result: None,
+                    logs: None,
+                    flow_status: None,
+                    deleted: uj.deleted,
+                    canceled: uj.canceled,
+                    canceled_by: uj.canceled_by,
+                    raw_code: None,
+                    canceled_reason: None,
+                    job_kind: uj.job_kind,
+                    schedule_path: uj.schedule_path,
+                    permissioned_as: uj.permissioned_as,
+                    raw_flow: None,
+                    is_flow_step: uj.is_flow_step,
+                    language: uj.language,
+                    is_skipped: uj.is_skipped,
+                    email: uj.email,
+                    visible_to_owner: uj.visible_to_owner,
+                    mem_peak: uj.mem_peak,
+                    tag: uj.tag,
+                    priority: uj.priority,
+                    labels: uj.labels,
+            })),
+            "QueuedJob" => Job::QueuedJob(JobExtended::new(uj.self_wait_time_ms, uj.aggregate_wait_time_ms, QueuedJob {
+                    workspace_id: uj.workspace_id,
+                    id: uj.id,
+                    parent_job: uj.parent_job,
+                    created_by: uj.created_by,
+                    created_at: uj.created_at,
+                    started_at: uj.started_at,
+                    script_hash: uj.script_hash,
+                    script_path: uj.script_path,
+                    args: None,
+                    running: uj.running.unwrap(),
+                    scheduled_for: uj.scheduled_for.unwrap(),
+                    logs: None,
+                    flow_status: None,
+                    raw_code: None,
+                    raw_lock: None,
+                    canceled: uj.canceled,
+                    canceled_by: uj.canceled_by,
+                    canceled_reason: None,
+                    last_ping: None,
+                    job_kind: uj.job_kind,
+                    schedule_path: uj.schedule_path,
+                    permissioned_as: uj.permissioned_as,
+                    raw_flow: None,
+                    is_flow_step: uj.is_flow_step,
+                    language: uj.language,
+                    same_worker: false,
+                    pre_run_error: None,
+                    email: uj.email,
+                    visible_to_owner: uj.visible_to_owner,
+                    suspend: uj.suspend,
+                    mem_peak: uj.mem_peak,
+                    root_job: None,
+                    leaf_jobs: None,
+                    tag: uj.tag,
+                    concurrent_limit: uj.concurrent_limit,
+                    concurrency_time_window_s: uj.concurrency_time_window_s,
+                    timeout: None,
+                    flow_step_id: None,
+                    cache_ttl: None,
+                    priority: uj.priority,
+            })),
             t => panic!("job type {} not valid", t),
         }
     }
@@ -3079,7 +3136,7 @@ pub async fn run_workflow_as_code(
         i += 1;
     }
 
-    let job = get_queued_job(&job_id, &w_id, &db).await?;
+    let job = get_queued_job_ex(&db, &w_id, job_id, true, None).await?;
 
     if *CLOUD_HOSTED {
         tracing::info!("workflow_as_code_tracing id {i} ");
@@ -3087,6 +3144,7 @@ pub async fn run_workflow_as_code(
     }
 
     let job = not_found_if_none(job, "Queued Job", &job_id.to_string())?;
+    let JobExtended { inner: job, .. } = job;
     let (job_payload, tag, _delete_after_use, timeout) = match job.job_kind {
         JobKind::Preview => (
             JobPayload::Code(RawCode {
@@ -4275,18 +4333,26 @@ async fn run_flow_dependencies_job(
     wait_result
 }
 
+#[derive(Deserialize)]
+struct BatchRawScript {
+    content: String,
+    language: Option<ScriptLang>,
+    lock: Option<String>,
+}
+
 #[derive(Deserialize)]
 struct BatchInfo {
     kind: String,
     flow_value: Option<FlowValue>,
     path: Option<String>,
+    rawscript: Option<BatchRawScript>,
 }
 
 #[tracing::instrument(level = "trace", skip_all)]
 async fn add_batch_jobs(
     authed: ApiAuthed,
     Extension(db): Extension<DB>,
-    Extension(rsmq): Extension<Option<rsmq_async::MultiplexedRsmq>>,
+    Extension(_rsmq): Extension<Option<rsmq_async::MultiplexedRsmq>>,
     Path((w_id, n)): Path<(String, i32)>,
     Json(batch_info): Json<BatchInfo>,
 ) -> error::JsonResult<Vec<Uuid>> {
@@ -4302,6 +4368,10 @@ async fn add_batch_jobs(
         concurrent_limit,
         concurrent_time_window_s,
         timeout,
+        raw_code,
+        raw_lock,
+        raw_flow,
+        flow_status
     ) = match batch_info.kind.as_str() {
         "script" => {
             if let Some(path) = batch_info.path {
@@ -4329,6 +4399,10 @@ async fn add_batch_jobs(
                     concurrent_limit,
                     concurrency_time_window_s,
                     timeout,
+                    None,
+                    None,
+                    None,
+                    None,
                 )
             } else {
                 Err(anyhow::anyhow!(
@@ -4336,61 +4410,72 @@ async fn add_batch_jobs(
                 ))?
             }
         }
-        "flow" => {
-            let mut uuids: Vec<Uuid> = Vec::new();
-            let payload = if let Some(ref fv) = batch_info.flow_value {
-                JobPayload::RawFlow { value: fv.clone(), path: None, restarted_from: None }
-            } else {
-                if let Some(path) = batch_info.path.as_ref() {
-                    JobPayload::Flow {
-                        path: path.to_string(),
-                        dedicated_worker: None,
-                        apply_preprocessor: false,
-                    }
-                } else {
-                    Err(anyhow::anyhow!(
-                        "Path is required if no value is not provided"
-                    ))?
-                }
-            };
-            let mut tx = PushIsolationLevel::IsolatedRoot(db.clone(), rsmq);
-            for _ in 0..n {
-                let ehm = HashMap::new();
-                let (uuid, ntx) = push(
-                    &db,
-                    tx,
-                    &w_id,
-                    payload.clone(),
-                    PushArgs::from(&ehm),
-                    authed.display_username(),
-                    &authed.email,
-                    username_to_permissioned_as(&authed.username),
+        "rawscript" => {
+            if let Some(rawscript) = batch_info.rawscript {
+                (
+                    None,
+                    None,
+                    JobKind::Preview,
+                    rawscript.language,
                     None,
                     None,
                     None,
                     None,
                     None,
-                    false,
-                    false,
-                    None,
-                    true,
+                    Some(rawscript.content),
+                    rawscript.lock,
                     None,
                     None,
-                    None,
-                    None,
-                    Some(&authed.clone().into()),
                 )
-                .await?;
-                tx = PushIsolationLevel::Transaction(ntx);
-                uuids.push(uuid);
+            } else {
+                Err(anyhow::anyhow!(
+                    "rawscript is required for `rawscript` kind"
+                ))?
             }
-            match tx {
-                PushIsolationLevel::Transaction(tx) => {
-                    tx.commit().await?;
-                }
-                _ => (),
-            }
-            return Ok(Json(uuids));
+        }
+        "flow" => {
+            let (mut value, job_kind, path) = if let Some(value) = batch_info.flow_value {
+                (value, JobKind::FlowPreview, None)
+            } else if let Some(path) = batch_info.path {
+                let value_json = sqlx::query!(
+                    "SELECT flow_version.value AS \"value: sqlx::types::Json<Box<RawValue>>\" FROM flow 
+                    LEFT JOIN flow_version
+                        ON flow_version.id = flow.versions[array_upper(flow.versions, 1)]
+                    WHERE flow.path = $1 AND flow.workspace_id = $2",
+                    &path, &w_id
+                )
+                .fetch_optional(&db)
+                .await?
+                .ok_or_else(|| Error::InternalErr(format!("not found flow at path {:?}", path)))?;
+                let value =
+                    serde_json::from_str::<FlowValue>(value_json.value.get()).map_err(|err| {
+                        Error::InternalErr(format!(
+                            "could not convert json to flow for {path}: {err:?}"
+                        ))
+                    })?;
+                (value, JobKind::Flow, Some(path))
+            } else {
+                Err(anyhow::anyhow!(
+                    "Path is required if no value is not provided"
+                ))?
+            };
+            add_virtual_items_if_necessary(&mut value.modules);
+            let flow_status = FlowStatus::new(&value);
+            (
+                None,                             // script_hash
+                path,                             // script_path
+                job_kind,                         // job_kind
+                None,                             // language
+                None,                             // dedicated_worker
+                value.concurrency_key.clone(),    // custom_concurrency_key
+                value.concurrent_limit.clone(),   // concurrent_limit
+                value.concurrency_time_window_s,  // concurrency_time_window_s
+                None,                             // timeout
+                None,                             // raw_code
+                None,                             // raw_lock
+                Some(value),                      // raw_flow
+                Some(flow_status),                // flow_status
+            )
         }
         "noop" => (
             None,
@@ -4402,6 +4487,10 @@ async fn add_batch_jobs(
             None,
             None,
             None,
+            None,
+            None,
+            None,
+            None,
         ),
         _ => {
             return Err(error::Error::BadRequest(format!(
@@ -4428,8 +4517,8 @@ async fn add_batch_jobs(
             select gen_random_uuid() as uuid from generate_series(1, $11)
         )
         INSERT INTO queue 
-            (id, script_hash, script_path, job_kind, language, args, tag, created_by, permissioned_as, email, scheduled_for, workspace_id, concurrent_limit, concurrency_time_window_s, timeout)
-            (SELECT uuid, $1, $2, $3, $4, ('{ "uuid": "' || uuid || '" }')::jsonb, $5, $6, $7, $8, $9, $10, $12, $13, $14 FROM uuid_table) 
+            (id, script_hash, script_path, job_kind, language, args, tag, created_by, permissioned_as, email, scheduled_for, workspace_id, concurrent_limit, concurrency_time_window_s, timeout, raw_code, raw_lock, raw_flow, flow_status)
+            (SELECT uuid, $1, $2, $3, $4, ('{ "uuid": "' || uuid || '" }')::jsonb, $5, $6, $7, $8, $9, $10, $12, $13, $14, $15, $16, $17, $18 FROM uuid_table) 
         RETURNING id"#,
             hash.map(|h| h.0),
             path,
@@ -4444,7 +4533,11 @@ async fn add_batch_jobs(
             n,
             concurrent_limit,
             concurrent_time_window_s,
-            timeout
+            timeout,
+            raw_code,
+            raw_lock,
+            raw_flow.map(sqlx::types::Json) as Option<sqlx::types::Json<FlowValue>>,
+            flow_status.map(sqlx::types::Json) as Option<sqlx::types::Json<FlowStatus>>
         )
         .fetch_all(&db)
         .await?;
@@ -5023,24 +5116,10 @@ async fn get_completed_job<'a>(
     Extension(db): Extension<DB>,
     Path((w_id, id)): Path<(String, Uuid)>,
 ) -> error::Result<Response> {
-    let job_o = sqlx::query_as::<_, CompletedJob>("SELECT id, workspace_id, parent_job, created_by, created_at, duration_ms, success, script_hash, script_path, 
-    CASE WHEN args is null or pg_column_size(args) < 90000 THEN args ELSE '\"WINDMILL_TOO_BIG\"'::jsonb END as args, CASE WHEN result is null or pg_column_size(result) < 90000 THEN result ELSE '\"WINDMILL_TOO_BIG\"'::jsonb END as result, logs, deleted, raw_code, canceled, canceled_by, canceled_reason, job_kind,
-    schedule_path, permissioned_as, flow_status, raw_flow, is_flow_step, language, started_at, is_skipped,
-    raw_lock, email, visible_to_owner, mem_peak, tag, priority, result->'wm_labels' as labels FROM completed_job WHERE id = $1 AND workspace_id = $2")
-        .bind(id)
-        .bind(&w_id)
-        .fetch_optional(&db)
+    let job_o = get_completed_job_ex(&db, &w_id, id, false, Some(&opt_authed))
         .await?;
 
     let cj = not_found_if_none(job_o, "Completed Job", id.to_string())?;
-
-    if opt_authed.is_none() && cj.created_by != "anonymous" {
-        return Err(Error::BadRequest(
-            "As a non logged in user, you can only see jobs ran by anonymous users".to_string(),
-        ));
-    }
-    let cj = format_completed_job_result(cj);
-
     let response = Json(cj).into_response();
     // let extra_log = query_scalar!(
     //     "SELECT substr(logs, $1) as logs FROM large_logs WHERE workspace_id = $2 AND job_id = $3",

backend/windmill-api/src/lib.rs

@@ -25,7 +25,6 @@ use argon2::Argon2;
 use axum::extract::DefaultBodyLimit;
 use axum::{middleware::from_extractor, routing::get, Extension, Router};
 use db::DB;
-use git_version::git_version;
 use http::HeaderValue;
 use reqwest::Client;
 use std::collections::HashMap;
@@ -40,7 +39,7 @@ use tower_http::{
 };
 use windmill_common::db::UserDB;
 use windmill_common::worker::{ALL_TAGS, CLOUD_HOSTED};
-use windmill_common::{BASE_URL, INSTANCE_NAME};
+use windmill_common::{BASE_URL, INSTANCE_NAME, utils::GIT_VERSION};
 
 use crate::scim_ee::has_scim_token;
 use windmill_common::error::AppError;
@@ -63,13 +62,14 @@ mod http_triggers;
 mod indexer_ee;
 mod inputs;
 mod integration;
+mod ai;
+
 #[cfg(feature = "parquet")]
 mod job_helpers_ee;
 pub mod job_metrics;
 pub mod jobs;
 pub mod oauth2_ee;
 mod oidc_ee;
-mod openai;
 mod raw_apps;
 mod resources;
 mod saml_ee;
@@ -93,11 +93,9 @@ mod workers;
 mod workspaces;
 mod workspaces_ee;
 
-pub const GIT_VERSION: &str =
-    git_version!(args = ["--tag", "--always"], fallback = "unknown-version");
-
 pub const DEFAULT_BODY_LIMIT: usize = 2097152 * 100; // 200MB
 
+
 lazy_static::lazy_static! {
 
     pub static ref REQUEST_SIZE_LIMIT: Arc<RwLock<usize>> = Arc::new(RwLock::new(DEFAULT_BODY_LIMIT));
@@ -282,7 +280,7 @@ pub async fn run_server(
                         .nest("/job_helpers", job_helpers_service)
                         .nest("/jobs", jobs::workspaced_service())
                         .nest("/oauth", oauth2_ee::workspaced_service())
-                        .nest("/openai", openai::workspaced_service())
+                        .nest("/ai", ai::workspaced_service())
                         .nest("/raw_apps", raw_apps::workspaced_service())
                         .nest("/resources", resources::workspaced_service())
                         .nest("/schedules", schedule::workspaced_service())

backend/windmill-api/src/openai.rs

@@ -1,344 +0,0 @@
-use std::collections::HashMap;
-
-use crate::db::{ApiAuthed, DB};
-
-use axum::{
-    body::Bytes,
-    extract::{Extension, Path, Query},
-    response::IntoResponse,
-    routing::post,
-    Router,
-};
-use quick_cache::sync::Cache;
-use reqwest::Client;
-use serde_json::value::RawValue;
-use windmill_audit::audit_ee::audit_log;
-use windmill_audit::ActionKind;
-use windmill_common::{
-    error::{to_anyhow, Error},
-    variables::build_crypt,
-};
-
-use crate::variables::decrypt;
-use serde::Deserialize;
-
-lazy_static::lazy_static! {
-    static ref HTTP_CLIENT: Client = reqwest::ClientBuilder::new()
-        .timeout(std::time::Duration::from_secs(60 * 5))
-        .user_agent("windmill/beta")
-        .build().unwrap();
-}
-
-pub fn workspaced_service() -> Router {
-    let router = Router::new().route("/proxy/*openai_path", post(proxy));
-
-    router
-}
-
-#[derive(Deserialize)]
-struct OpenaiResource {
-    api_key: String,
-    organization_id: Option<String>,
-}
-
-#[derive(Deserialize)]
-struct OpenaiClientCredentialsOauthResource {
-    client_id: String,
-    client_secret: String,
-    token_url: String,
-    user: Option<String>,
-}
-
-#[derive(Deserialize)]
-#[serde(untagged)]
-enum OpenaiConfig {
-    Resource(OpenaiResource),
-    ClientCredentialsOauthResource(OpenaiClientCredentialsOauthResource),
-}
-
-struct Variable {
-    value: String,
-    is_secret: bool,
-}
-async fn get_variable_or_self(path: String, db: &DB, w_id: &String) -> Result<String, Error> {
-    if !path.starts_with("$var:") {
-        return Ok(path);
-    }
-    let path = path.strip_prefix("$var:").unwrap().to_string();
-    let mut variable = sqlx::query_as!(
-        Variable,
-        "SELECT value, is_secret
-        FROM variable
-        WHERE path = $1 AND workspace_id = $2",
-        &path,
-        &w_id
-    )
-    .fetch_one(db)
-    .await?;
-    if variable.is_secret {
-        let mc = build_crypt(&db, &w_id).await?;
-        variable.value = decrypt(&mc, variable.value)?;
-    }
-    Ok(variable.value)
-}
-
-lazy_static::lazy_static! {
-    pub static ref OPENAI_AZURE_BASE_PATH: Option<String> = std::env::var("OPENAI_AZURE_BASE_PATH").ok();
-    static ref OPENAI_KEY_CACHE: Cache<String, OpenaiKeyCache> = Cache::new(500);
-}
-
-#[derive(Deserialize)]
-struct OpenaiCredentials {
-    access_token: String,
-}
-async fn get_openai_key_using_credentials_flow(
-    mut resource: OpenaiClientCredentialsOauthResource,
-    db: &DB,
-    w_id: &String,
-) -> Result<String, Error> {
-    resource.client_id = get_variable_or_self(resource.client_id, &db, &w_id).await?;
-    resource.client_secret = get_variable_or_self(resource.client_secret, &db, &w_id).await?;
-    resource.token_url = get_variable_or_self(resource.token_url, &db, &w_id).await?;
-    let mut params = HashMap::new();
-    params.insert("grant_type", "client_credentials");
-    let response = HTTP_CLIENT
-        .post(resource.token_url)
-        .form(&params)
-        .basic_auth(resource.client_id, Some(resource.client_secret))
-        .send()
-        .await
-        .map_err(|err| {
-            Error::InternalErr(format!(
-                "Failed to get OpenAI credentials using credentials flow: {}",
-                err
-            ))
-        })?;
-    let response = response.json::<OpenaiCredentials>().await.map_err(|err| {
-        Error::InternalErr(format!(
-            "Failed to parse OpenAI credentials from credentials flow: {}",
-            err
-        ))
-    })?;
-    Ok(response.access_token)
-}
-
-#[derive(Clone)]
-struct OpenaiKeyCache {
-    api_key: String,
-    organization_id: Option<String>,
-    azure_base_path: Option<String>,
-    user: Option<String>,
-    expires_at: std::time::Instant,
-}
-
-impl OpenaiKeyCache {
-    fn new(
-        api_key: String,
-        organization_id: Option<String>,
-        azure_base_path: Option<String>,
-        expires_at: std::time::Instant,
-        user: Option<String>,
-    ) -> Self {
-        Self { api_key, organization_id, azure_base_path, expires_at, user }
-    }
-    fn is_expired(&self) -> bool {
-        self.expires_at < std::time::Instant::now()
-    }
-}
-
-#[derive(Deserialize)]
-struct ProxyQueryParams {
-    no_cache: Option<bool>,
-}
-async fn proxy(
-    authed: ApiAuthed,
-    Extension(db): Extension<DB>,
-    Path((w_id, openai_path)): Path<(String, String)>,
-    Query(query_params): Query<ProxyQueryParams>,
-    mut body: Bytes,
-) -> impl IntoResponse {
-    let workspace_cache = OPENAI_KEY_CACHE.get(&w_id);
-    let (api_key, organization_id, azure_base_path, user) = if query_params
-        .no_cache
-        .unwrap_or(false)
-        || workspace_cache.is_none()
-        || workspace_cache.clone().unwrap().is_expired()
-    {
-        let openai_resource_path = sqlx::query_scalar!(
-            "SELECT openai_resource_path FROM workspace_settings WHERE workspace_id = $1",
-            &w_id
-        )
-        .fetch_one(&db)
-        .await?;
-
-        if openai_resource_path.is_none() {
-            return Err(Error::InternalErr(
-                "OpenAI resource not configured".to_string(),
-            ));
-        }
-
-        let openai_resource_path = openai_resource_path.unwrap();
-
-        let resource = sqlx::query_scalar!(
-            "SELECT value
-            FROM resource
-            WHERE path = $1 AND workspace_id = $2",
-            &openai_resource_path,
-            &w_id
-        )
-        .fetch_optional(&db)
-        .await?
-        .ok_or_else(|| {
-            Error::InternalErr(format!(
-                "Could not find the OpenAI resource at path {openai_resource_path}, update the resource path in the workspace settings"
-            ))
-        })?;
-
-        if resource.is_none() {
-            return Err(Error::InternalErr(
-                "OpenAI resource missing value".to_string(),
-            ));
-        }
-
-        let config: OpenaiConfig = serde_json::from_value(resource.unwrap())
-            .map_err(|e| Error::InternalErr(format!("validating openai resource {e:#}")))?;
-
-        let mut user = None::<String>;
-        let mut resource = match config {
-            OpenaiConfig::Resource(resource) => {
-                tracing::debug!("Getting OpenAI key from static resource");
-                resource
-            }
-            OpenaiConfig::ClientCredentialsOauthResource(resource) => {
-                tracing::debug!("Getting OpenAI key with client credentials flow");
-                user = resource.user.clone();
-                let token = get_openai_key_using_credentials_flow(resource, &db, &w_id).await?;
-                OpenaiResource { api_key: token, organization_id: None }
-            }
-        };
-
-        resource.api_key = get_variable_or_self(resource.api_key, &db, &w_id).await?;
-
-        if resource.organization_id.is_some() {
-            resource.organization_id =
-                Some(get_variable_or_self(resource.organization_id.unwrap(), &db, &w_id).await?);
-        }
-
-        if user.is_some() {
-            user = Some(get_variable_or_self(user.unwrap(), &db, &w_id).await?);
-        }
-
-        let expires_at = std::time::Instant::now() + std::time::Duration::from_secs(60);
-
-        let azure_base_path = sqlx::query_scalar!(
-            "SELECT value
-            FROM global_settings
-            WHERE name = 'openai_azure_base_path'",
-        )
-        .fetch_optional(&db)
-        .await?;
-
-        let azure_base_path = if let Some(azure_base_path) = azure_base_path {
-            Some(
-                serde_json::from_value::<String>(azure_base_path).map_err(|e| {
-                    Error::InternalErr(format!("validating openai azure base path {e:#}"))
-                })?,
-            )
-        } else {
-            OPENAI_AZURE_BASE_PATH.clone()
-        };
-
-        let workspace_cache = OpenaiKeyCache::new(
-            resource.api_key.clone(),
-            resource.organization_id.clone(),
-            azure_base_path.clone(),
-            expires_at,
-            user.clone(),
-        );
-        OPENAI_KEY_CACHE.insert(w_id.clone(), workspace_cache);
-        (
-            resource.api_key,
-            resource.organization_id,
-            azure_base_path,
-            user,
-        )
-    } else {
-        tracing::debug!("Using cached OpenAI key");
-        let workspace_cache = workspace_cache.unwrap();
-        (
-            workspace_cache.api_key.clone(),
-            workspace_cache.organization_id.clone(),
-            workspace_cache.azure_base_path.clone(),
-            workspace_cache.user.clone(),
-        )
-    };
-
-    if user.is_some() {
-        tracing::debug!("Adding user to request body");
-        let mut json_body: HashMap<String, Box<RawValue>> = serde_json::from_slice(&body)
-            .map_err(|e| Error::InternalErr(format!("Failed to parse request body: {}", e)))?;
-
-        let user_json_string = serde_json::Value::String(user.unwrap()).to_string(); // makes sure to escape characters
-
-        json_body.insert(
-            "user".to_string(),
-            RawValue::from_string(user_json_string)
-                .map_err(|e| Error::InternalErr(format!("Failed to parse user: {}", e)))?,
-        );
-
-        body = serde_json::to_vec(&json_body)
-            .map_err(|e| Error::InternalErr(format!("Failed to reserialize request body: {}", e)))?
-            .into();
-    }
-
-    let base_url = if let Some(base_url) = azure_base_path {
-        base_url
-    } else {
-        "https://api.openai.com/v1".to_string()
-    };
-
-    let url = format!("{}/{}", base_url, openai_path);
-    let mut request = HTTP_CLIENT
-        .post(url)
-        .header("content-type", "application/json")
-        .body(body);
-
-    if base_url != "https://api.openai.com/v1" {
-        request = request
-            .header("api-key", api_key)
-            .query(&[("api-version", "2023-05-15")])
-    } else {
-        request = request.header("authorization", format!("Bearer {}", api_key))
-    }
-
-    if let Some(org_id) = organization_id {
-        request = request.header("OpenAI-Organization", org_id);
-    }
-
-    let response = request.send().await.map_err(to_anyhow)?;
-
-    let mut tx = db.begin().await?;
-    audit_log(
-        &mut *tx,
-        &authed,
-        "openai.request",
-        ActionKind::Execute,
-        &w_id,
-        Some(&authed.email),
-        Some([("openai_path", &format!("{:?}", openai_path)[..])].into()),
-    )
-    .await?;
-    tx.commit().await?;
-
-    if response.error_for_status_ref().is_err() {
-        return Err(Error::OpenAIError(
-            response.text().await.unwrap_or("".to_string()),
-        ));
-    }
-
-    let status_code = response.status();
-    let headers = response.headers().clone();
-    let stream = response.bytes_stream();
-
-    Ok((status_code, headers, axum::body::Body::from_stream(stream)))
-}

backend/windmill-api/src/resources.rs

@@ -300,7 +300,7 @@ async fn get_resource(
     .await?;
     tx.commit().await?;
     if resource_o.is_none() {
-        explain_resource_perm_error(&path, &w_id, &db).await?;
+        explain_resource_perm_error(&path, &w_id, &db, &authed).await?;
     }
     let resource = not_found_if_none(resource_o, "Resource", path)?;
     Ok(Json(resource))
@@ -343,7 +343,7 @@ async fn get_resource_value(
 
     tx.commit().await?;
     if value_o.is_none() {
-        explain_resource_perm_error(&path, &w_id, &db).await?;
+        explain_resource_perm_error(&path, &w_id, &db, &authed).await?;
     }
 
     let value = not_found_if_none(value_o, "Resource", path)?;
@@ -354,6 +354,7 @@ async fn explain_resource_perm_error(
     path: &str,
     w_id: &str,
     db: &sqlx::Pool<Postgres>,
+    authed: &ApiAuthed,
 ) -> windmill_common::error::Result<()> {
     let extra_perms = sqlx::query_scalar!(
         "SELECT extra_perms from resource WHERE path = $1 AND workspace_id = $2",
@@ -378,12 +379,12 @@ async fn explain_resource_perm_error(
         .fetch_optional(db)
         .await?;
         return Err(Error::NotAuthorized(format!(
-            "Resource exists but you don't have access to it:\nresource perms: {}\nfolder perms: {}",
+            "Resource exists but you don't have access to it:\nresource perms: {}\nfolder perms: {}\nauthed as: {authed:?}",
             serde_json::to_string_pretty(&extra_perms).unwrap_or_default(), serde_json::to_string_pretty(&folder_extra_perms).unwrap_or_default()
         )));
     } else {
         return Err(Error::NotAuthorized(format!(
-            "Resource exists but you don't have access to it:\nresource perms: {}",
+            "Resource exists but you don't have access to it:\nresource perms: {}\nauthed as: {authed:?}",
             serde_json::to_string_pretty(&extra_perms).unwrap_or_default()
         )));
     }
@@ -457,7 +458,7 @@ pub async fn get_resource_value_interpolated_internal(
     .await?;
     tx.commit().await?;
     if value_o.is_none() {
-        explain_resource_perm_error(path, workspace, db).await?;
+        explain_resource_perm_error(path, workspace, db, &authed).await?;
     }
 
     let value = not_found_if_none(value_o, "Resource", path)?;

backend/windmill-api/src/settings.rs

@@ -58,7 +58,10 @@ pub fn global_service() -> Router {
         )
         .route("/renew_license_key", post(renew_license_key))
         .route("/customer_portal", post(create_customer_portal_session))
-        .route("/test_critical_channels", post(test_critical_channels));
+        .route("/test_critical_channels", post(test_critical_channels))
+        .route("/critical_alerts", get(get_critical_alerts))
+        .route("/critical_alerts/:id/acknowledge", post(acknowledge_critical_alert))
+        .route("/critical_alerts/acknowledge_all", post(acknowledge_all_critical_alerts));
 
     #[cfg(feature = "parquet")]
     {
@@ -430,3 +433,50 @@ pub async fn test_critical_channels(
 pub async fn test_critical_channels() -> Result<String> {
     Ok("Critical channels require EE".to_string())
 }
+
+#[cfg(feature = "enterprise")]
+pub async fn get_critical_alerts(
+    Extension(db): Extension<DB>,
+    authed: ApiAuthed,
+    Query(params): Query<crate::utils::AlertQueryParams>,
+) -> JsonResult<Vec<crate::utils::CriticalAlert>> {
+    require_super_admin(&db, &authed.email).await?;
+
+    crate::utils::get_critical_alerts(db, params, None).await
+}
+
+#[cfg(not(feature = "enterprise"))]
+pub async fn get_critical_alerts() -> error::Error {
+    error::Error::NotFound("Critical Alerts require EE".to_string())
+}
+
+#[cfg(feature = "enterprise")]
+pub async fn acknowledge_critical_alert(
+    Extension(db): Extension<DB>,
+    authed: ApiAuthed,
+    Path(id): Path<i32>,
+) -> error::Result<String> {
+    require_super_admin(&db, &authed.email).await?;
+    
+    crate::utils::acknowledge_critical_alert(db, None, id).await
+}
+
+#[cfg(not(feature = "enterprise"))]
+pub async fn acknowledge_critical_alert() -> error::Error {
+    error::Error::NotFound("Critical Alerts require EE".to_string())
+}
+
+#[cfg(feature = "enterprise")]
+pub async fn acknowledge_all_critical_alerts(
+    Extension(db): Extension<DB>,
+    authed: ApiAuthed,
+) -> error::Result<String> {
+    require_super_admin(&db, &authed.email).await?;
+
+    crate::utils::acknowledge_all_critical_alerts(db, None).await
+}
+
+#[cfg(not(feature = "enterprise"))]
+pub async fn acknowledge_all_critical_alerts() -> error::Error {
+    error::Error::NotFound("Critical Alerts require EE".to_string())
+}

backend/windmill-api/src/users.rs

@@ -90,6 +90,9 @@ pub fn global_service() -> Router {
         .route("/accept_invite", post(accept_invite))
         .route("/list_as_super_admin", get(list_users_as_super_admin))
         .route("/setpassword", post(set_password))
+        .route("/set_password_of/:user", post(set_password_of_user))
+        .route("/set_login_type/:user", post(set_login_type))
+
         .route("/create", post(create_user))
         .route("/update/:user", post(update_user))
         .route("/delete/:user", delete(delete_user))
@@ -746,10 +749,20 @@ pub async fn fetch_api_authed(
     username_override: String,
 ) -> error::Result<ApiAuthed> {
     let permissioned_as = username_to_permissioned_as(username.as_str());
+    fetch_api_authed_from_permissioned_as(permissioned_as, email, w_id, db, username_override).await
+}
+
+pub async fn fetch_api_authed_from_permissioned_as(
+    permissioned_as: String,
+    email: String,
+    w_id: &str,
+    db: &DB,
+    username_override: String,
+) -> error::Result<ApiAuthed> {
     let authed =
         fetch_authed_from_permissioned_as(permissioned_as, email.clone(), w_id, db).await?;
     Ok(ApiAuthed {
-        username: username,
+        username: authed.username,
         email: email,
         is_admin: authed.is_admin,
         is_operator: authed.is_operator,
@@ -856,6 +869,12 @@ pub struct EditPassword {
     pub password: String,
 }
 
+#[derive(Deserialize)]
+pub struct EditLoginType {
+    pub login_type: String,
+}
+
+
 #[derive(FromRow, Serialize)]
 pub struct TruncatedToken {
     pub label: Option<String>,
@@ -2018,7 +2037,52 @@ async fn set_password(
     authed: ApiAuthed,
     Json(ep): Json<EditPassword>,
 ) -> Result<String> {
-    crate::users_ee::set_password(db, argon2, authed, ep).await
+    let email = authed.email.clone();
+    crate::users_ee::set_password(db, argon2, authed, &email, ep).await
+}
+
+async fn set_password_of_user(
+    Extension(db): Extension<DB>,
+    Extension(argon2): Extension<Arc<Argon2<'_>>>,
+    Path(email): Path<String>,
+    authed: ApiAuthed,
+    Json(ep): Json<EditPassword>,
+) -> Result<String> {
+    require_super_admin(&db, &authed.email).await?;
+    crate::users_ee::set_password(db, argon2, authed, &email, ep).await
+}
+
+async fn set_login_type(
+    Extension(db): Extension<DB>,
+    Path(email): Path<String>,
+    authed: ApiAuthed,
+    Json(et): Json<EditLoginType>,
+) -> Result<String> {
+    require_super_admin(&db, &authed.email).await?;
+    let mut tx = db.begin().await?;
+
+    sqlx::query!(
+        "UPDATE password SET login_type = $1 WHERE email = $2",
+        et.login_type,
+        email
+    )
+    .execute(&mut *tx)
+    .await?;
+
+    audit_log(
+        &mut *tx,
+        &authed,
+        "users.set_login_type",
+        ActionKind::Update,
+        "global",
+        Some(&email),
+        None,
+    )
+    .await?;
+
+    tx.commit().await?;
+    Ok(format!("login type of {} updated to {}", email, et.login_type))
+    
 }
 
 async fn login(

backend/windmill-api/src/users_ee.rs

@@ -27,6 +27,7 @@ pub async fn set_password(
     _db: DB,
     _argon2: Arc<Argon2<'_>>,
     _authed: ApiAuthed,
+    _user_email: &str,
     _ep: EditPassword,
 ) -> Result<String> {
     Err(Error::InternalErr(

backend/windmill-api/src/utils.rs

@@ -16,6 +16,12 @@ use windmill_common::{
     DB,
 };
 
+#[cfg(feature = "enterprise")]
+use windmill_common::error::JsonResult;
+
+#[cfg(feature = "enterprise")]
+use axum::Json;
+
 #[derive(Deserialize)]
 pub struct WithStarredInfoQuery {
     pub with_starred_info: Option<bool>,
@@ -162,3 +168,159 @@ pub fn content_plain(body: Body) -> Response {
         .body(body)
         .unwrap()
 }
+
+use serde::Serialize;
+
+#[derive(Serialize)]
+pub struct CriticalAlert {
+    id: i32,
+    alert_type: String,
+    message: String,
+    created_at: chrono::DateTime<chrono::Utc>,
+    acknowledged: Option<bool>,
+    workspace_id: Option<String>,
+}
+
+#[cfg(feature = "enterprise")]
+#[derive(Deserialize, Debug)]
+pub struct AlertQueryParams {
+    pub page: Option<i32>,
+    pub page_size: Option<i32>,
+    pub acknowledged: Option<bool>,
+}
+
+#[cfg(feature = "enterprise")]
+pub async fn get_critical_alerts(
+    db: DB,
+    params: AlertQueryParams,
+    workspace_id: Option<String>,
+) -> JsonResult<Vec<CriticalAlert>> {
+    let page = params.page.unwrap_or(1).max(1);
+    let page_size = params.page_size.unwrap_or(10).min(100) as i64;
+    let offset = ((page - 1) * page_size as i32) as i64;
+
+    let alerts = if let Some(workspace_id) = workspace_id {
+        // `workspace_id` is provided => workspace admin
+        if params.acknowledged.is_none() {
+            // Case: return all rows where `workspace_id` matches
+            sqlx::query_as!(
+                CriticalAlert,
+                "SELECT id, alert_type, message, created_at, COALESCE(acknowledged_workspace, false) AS acknowledged, workspace_id
+                 FROM alerts
+                 WHERE workspace_id = $1
+                 ORDER BY created_at DESC
+                 LIMIT $2 OFFSET $3",
+                workspace_id,
+                page_size,
+                offset
+            )
+            .fetch_all(&db)
+            .await?
+        } else {
+            // Case: return rows where `acknowledged_workspace` matches `params.acknowledged`
+            sqlx::query_as!(
+                CriticalAlert,
+                "SELECT id, alert_type, message, created_at, COALESCE(acknowledged_workspace, false) AS acknowledged, workspace_id
+                 FROM alerts
+                 WHERE workspace_id = $1 AND COALESCE(acknowledged_workspace, false) = $2
+                 ORDER BY created_at DESC
+                 LIMIT $3 OFFSET $4",
+                workspace_id,
+                params.acknowledged,
+                page_size,
+                offset
+            )
+            .fetch_all(&db)
+            .await?
+        }
+    } else {
+        // `workspace_id` is not provided => superadmin
+        if params.acknowledged.is_none() {
+            // Case: Return all rows unfiltered with global acknowledged as acknowledged
+            sqlx::query_as!(
+                CriticalAlert,
+                "SELECT id, alert_type, message, created_at, COALESCE(acknowledged, false) AS acknowledged, workspace_id
+                 FROM alerts
+                 ORDER BY created_at DESC
+                 LIMIT $1 OFFSET $2",
+                page_size,
+                offset
+            )
+            .fetch_all(&db)
+            .await?
+        } else {
+            // Case: Return rows where global acknowledged matches params.acknowledged
+            sqlx::query_as!(
+                CriticalAlert,
+                "SELECT id, alert_type, message, created_at, COALESCE(acknowledged, false) AS acknowledged, workspace_id
+                 FROM alerts
+                 WHERE COALESCE(acknowledged, false) = $1
+                 ORDER BY created_at DESC
+                 LIMIT $2 OFFSET $3",
+                params.acknowledged,
+                page_size,
+                offset
+            )
+            .fetch_all(&db)
+            .await?
+        }
+    };
+
+    Ok(Json(alerts))
+}
+
+#[cfg(feature = "enterprise")]
+pub async fn acknowledge_critical_alert(
+    db: DB,
+    workspace_id: Option<String>,
+    id: i32,
+) -> error::Result<String> {
+    sqlx::query!(
+        "UPDATE alerts
+         SET
+           acknowledged = true,
+           acknowledged_workspace = CASE
+             WHEN $2::text IS NOT NULL AND workspace_id = $2 THEN true
+             ELSE acknowledged_workspace
+           END
+         WHERE id = $1",
+        id,
+        workspace_id
+    )
+    .execute(&db)
+    .await?;
+
+    tracing::info!(
+        "Acknowledged critical alert with id: {}{}",
+        id,
+        workspace_id.map_or_else(|| "".to_string(), |w| format!(" for workspace_id: {}", w))
+    );
+    Ok("Critical alert acknowledged".to_string())
+}
+
+#[cfg(feature = "enterprise")]
+pub async fn acknowledge_all_critical_alerts(
+    db: DB,
+    workspace_id: Option<String>,
+) -> error::Result<String> {
+    sqlx::query!(
+        "UPDATE alerts 
+         SET
+           acknowledged = true,
+           acknowledged_workspace = CASE
+             WHEN $1::text IS NOT NULL THEN true
+             ELSE acknowledged_workspace
+           END
+         WHERE ($1::text IS NOT NULL AND workspace_id = $1)
+            OR ($1::text IS NULL)",
+        workspace_id
+    )
+    .execute(&db)
+    .await?;
+
+    tracing::info!(
+        "Acknowledged all unacknowledged critical alerts{}",
+        workspace_id.map_or_else(|| "".to_string(), |w| format!(" for workspace_id: {}", w))
+    );
+    Ok("All unacknowledged critical alerts acknowledged".to_string())
+}

backend/windmill-api/src/websocket_triggers.rs

@@ -754,12 +754,14 @@ async fn disable_with_error(db: &DB, ws_trigger: &WebsocketTrigger, error: Strin
     )
     .execute(db).await {
         Ok(_) => {
-            report_critical_error(format!("Disabling websocket {} because of error: {}", ws_trigger.url, error), db.clone()).await;
+            report_critical_error(format!("Disabling websocket {} because of error: {}", ws_trigger.url, error), db.clone(), Some(&ws_trigger.workspace_id), None).await;
         },
         Err(disable_err) => {
             report_critical_error(
                 format!("Could not disable websocket {} with err {}, disabling because of error {}", ws_trigger.path, disable_err, error), 
-                db.clone()
+                db.clone(),
+                Some(&ws_trigger.workspace_id),
+                None,
             ).await;
         }
     }
@@ -896,7 +898,7 @@ async fn listen_to_websocket(
                                                                 }
                                                                 if should_handle {
                                                                     if let Err(err) = run_job(&db, rsmq.clone(), &ws_trigger, text).await {
-                                                                        report_critical_error(format!("Failed to trigger job from websocket {}: {:?}", ws_trigger.url, err), db.clone()).await;
+                                                                        report_critical_error(format!("Failed to trigger job from websocket {}: {:?}", ws_trigger.url, err), db.clone(), Some(&ws_trigger.workspace_id), None).await;
                                                                     };
                                                                 }
                                                             },

backend/windmill-api/src/workers.rs

@@ -36,6 +36,7 @@ pub fn global_service() -> Router {
         )
         .route("/get_default_tags", get(get_default_tags))
         .route("/queue_metrics", get(get_queue_metrics))
+        .route("/queue_counts", get(get_queue_counts))
 }
 
 #[derive(FromRow, Serialize, Deserialize)]
@@ -176,3 +177,12 @@ async fn get_queue_metrics(
 
     Ok(Json(queue_metrics))
 }
+
+async fn get_queue_counts(
+    authed: ApiAuthed,
+    Extension(db): Extension<DB>,
+) -> JsonResult<std::collections::HashMap<String, u32>> {
+    require_super_admin(&db, &authed.email).await?;
+    let queue_counts = windmill_common::queue::get_queue_counts(&db).await;
+    Ok(Json(queue_counts))
+}

backend/windmill-api/src/workspaces.rs

@@ -8,6 +8,7 @@
 
 use std::collections::HashMap;
 
+use crate::ai::{AiResource, AI_KEY_CACHE};
 use crate::db::ApiAuthed;
 use crate::users_ee::send_email_if_possible;
 use crate::utils::get_instance_username_or_create_pending;
@@ -118,7 +119,11 @@ pub fn workspaced_service() -> Router {
         .route("/change_workspace_name", post(change_workspace_name))
         .route("/change_workspace_id", post(change_workspace_id))
         .route("/usage", get(get_usage))
-        .route("/used_triggers", get(get_used_triggers));
+        .route("/used_triggers", get(get_used_triggers))
+        .route("/critical_alerts", get(get_critical_alerts))
+        .route("/critical_alerts/:id/acknowledge", post(acknowledge_critical_alert))
+        .route("/critical_alerts/acknowledge_all", post(acknowledge_all_critical_alerts))
+        .route("/critical_alerts/mute", post(mute_critical_alerts));
 
     #[cfg(feature = "stripe")]
     {
@@ -168,7 +173,7 @@ pub struct WorkspaceSettings {
     pub plan: Option<String>,
     pub webhook: Option<String>,
     pub deploy_to: Option<String>,
-    pub openai_resource_path: Option<String>,
+    pub ai_resource: Option<serde_json::Value>,
     pub code_completion_enabled: bool,
     pub error_handler: Option<String>,
     pub error_handler_extra_args: Option<serde_json::Value>,
@@ -179,6 +184,7 @@ pub struct WorkspaceSettings {
     pub default_app: Option<String>,
     pub automatic_billing: bool,
     pub default_scripts: Option<serde_json::Value>,
+    pub mute_critical_alerts: Option<bool>,
 }
 
 #[derive(FromRow, Serialize, Debug)]
@@ -234,7 +240,7 @@ struct EditWebhook {
 
 #[derive(Deserialize)]
 struct EditCopilotConfig {
-    openai_resource_path: Option<String>,
+    ai_resource: Option<serde_json::Value>,
     code_completion_enabled: bool,
 }
 
@@ -645,23 +651,33 @@ async fn edit_copilot_config(
 
     let mut tx = db.begin().await?;
 
-    if let Some(openai_resource_path) = &eo.openai_resource_path {
+    if let Some(ai_resource) = &eo.ai_resource {
+        let path = serde_json::from_value::<AiResource>(ai_resource.clone())
+            .map_err(|e| Error::BadRequest(e.to_string()))?
+            .path;
         sqlx::query!(
-            "UPDATE workspace_settings SET openai_resource_path = $1, code_completion_enabled = $2 WHERE workspace_id = $3",
-            openai_resource_path,
+            "UPDATE workspace_settings SET ai_resource = $1, code_completion_enabled = $2 WHERE workspace_id = $3",
+            ai_resource,
             eo.code_completion_enabled,
             &w_id
         )
         .execute(&mut *tx)
         .await?;
+
+        if let Some(cached) = AI_KEY_CACHE.get(&w_id) {
+            if cached.path != path {
+                AI_KEY_CACHE.remove(&w_id);
+            }
+        }
     } else {
         sqlx::query!(
-            "UPDATE workspace_settings SET openai_resource_path = NULL, code_completion_enabled = $1 WHERE workspace_id = $2",
+            "UPDATE workspace_settings SET ai_resource = NULL, code_completion_enabled = $1 WHERE workspace_id = $2",
             eo.code_completion_enabled,
             &w_id,
         )
         .execute(&mut *tx)
         .await?;
+        AI_KEY_CACHE.remove(&w_id);
     }
     audit_log(
         &mut *tx,
@@ -672,10 +688,7 @@ async fn edit_copilot_config(
         Some(&authed.email),
         Some(
             [
-                (
-                    "openai_resource_path",
-                    &format!("{:?}", eo.openai_resource_path)[..],
-                ),
+                ("ai_resource", &format!("{:?}", eo.ai_resource)[..]),
                 (
                     "code_completion_enabled",
                     &format!("{:?}", eo.code_completion_enabled)[..],
@@ -692,7 +705,8 @@ async fn edit_copilot_config(
 
 #[derive(Serialize)]
 struct CopilotInfo {
-    pub exists_openai_resource_path: bool,
+    pub ai_provider: String,
+    pub exists_ai_resource: bool,
     pub code_completion_enabled: bool,
 }
 async fn get_copilot_info(
@@ -701,16 +715,32 @@ async fn get_copilot_info(
 ) -> JsonResult<CopilotInfo> {
     let mut tx = db.begin().await?;
     let record = sqlx::query!(
-        "SELECT openai_resource_path, code_completion_enabled FROM workspace_settings WHERE workspace_id = $1",
+        "SELECT ai_resource, code_completion_enabled FROM workspace_settings WHERE workspace_id = $1",
         &w_id
     )
     .fetch_one(&mut *tx)
     .await
-    .map_err(|e| Error::InternalErr(format!("getting openai_resource_path and code_completion_enabled: {e:#}")))?;
+    .map_err(|e| Error::InternalErr(format!("getting ai_resource and code_completion_enabled: {e:#}")))?;
     tx.commit().await?;
 
+    let (ai_provider, exists_ai_resource) = if let Some(ai_resource) = record.ai_resource {
+        let ai_resource = serde_json::from_value::<AiResource>(ai_resource);
+        let exist = ai_resource.is_ok();
+        (
+            if exist {
+                ai_resource.unwrap().provider
+            } else {
+                "".to_string()
+            },
+            exist,
+        )
+    } else {
+        ("".to_string(), false)
+    };
+
     Ok(Json(CopilotInfo {
-        exists_openai_resource_path: record.openai_resource_path.is_some(),
+        ai_provider,
+        exists_ai_resource,
         code_completion_enabled: record.code_completion_enabled,
     }))
 }
@@ -2223,7 +2253,7 @@ struct SimplifiedSettings {
     error_handler: Option<String>,
     error_handler_extra_args: Option<Value>,
     error_handler_muted_on_cancel: bool,
-    openai_resource_path: Option<String>,
+    ai_resource: Option<serde_json::Value>,
     code_completion_enabled: bool,
     large_file_storage: Option<Value>,
     git_sync: Option<Value>,
@@ -2588,7 +2618,7 @@ async fn tarball_workspace(
                 webhook, 
                 deploy_to, 
                 error_handler, 
-                openai_resource_path, 
+                ai_resource, 
                 code_completion_enabled, 
                 error_handler_extra_args, 
                 error_handler_muted_on_cancel, 
@@ -3050,3 +3080,93 @@ async fn get_usage(Extension(db): Extension<DB>, Path(w_id): Path<String>) -> Re
     .unwrap_or(0);
     Ok(usage.to_string())
 }
+
+#[cfg(feature = "enterprise")]
+pub async fn get_critical_alerts(
+    Extension(db): Extension<DB>,
+    Path(w_id): Path<String>,
+    authed: ApiAuthed,
+    Query(params): Query<crate::utils::AlertQueryParams>,
+) -> JsonResult<Vec<crate::utils::CriticalAlert>> {
+    require_admin(authed.is_admin, &authed.username)?;
+
+    crate::utils::get_critical_alerts(db, params, Some(w_id)).await
+}
+
+#[cfg(not(feature = "enterprise"))]
+pub async fn get_critical_alerts() -> Error {
+    Error::NotFound("Critical Alerts require EE".to_string())
+}
+
+#[cfg(feature = "enterprise")]
+pub async fn acknowledge_critical_alert(
+    Extension(db): Extension<DB>,
+    Path((w_id, id)): Path<(String, i32)>,
+    authed: ApiAuthed,
+) -> Result<String> {
+    require_admin(authed.is_admin, &authed.username)?;
+    crate::utils::acknowledge_critical_alert(db, Some(w_id), id).await 
+}
+
+#[cfg(not(feature = "enterprise"))]
+pub async fn acknowledge_critical_alert() -> Error {
+    Error::NotFound("Critical Alerts require EE".to_string())
+}
+
+#[cfg(feature = "enterprise")]
+pub async fn acknowledge_all_critical_alerts(
+    Extension(db): Extension<DB>,
+    Path(w_id): Path<String>,
+    authed: ApiAuthed,
+) -> Result<String> {
+    require_admin(authed.is_admin, &authed.username)?;
+    crate::utils::acknowledge_all_critical_alerts(db, Some(w_id)).await
+}
+
+#[cfg(not(feature = "enterprise"))]
+pub async fn acknowledge_all_critical_alerts() -> Error {
+    Error::NotFound("Critical Alerts require EE".to_string())
+}
+
+
+#[cfg(feature = "enterprise")]
+#[derive(Deserialize)]
+pub struct MuteCriticalAlertRequest {
+    pub mute_critical_alerts: Option<bool>,
+}
+
+#[cfg(feature = "enterprise")]
+async fn mute_critical_alerts(
+    Extension(db): Extension<DB>,
+    Path(w_id): Path<String>,
+    ApiAuthed { is_admin, username, .. }: ApiAuthed,
+    Json(m_r): Json<MuteCriticalAlertRequest>,
+) -> Result<String> {
+    require_admin(is_admin, &username)?;
+
+    let mute_alerts = m_r.mute_critical_alerts.unwrap_or(false);
+
+    if mute_alerts {
+        sqlx::query!(
+            "UPDATE alerts SET acknowledged_workspace = true, acknowledged = true WHERE workspace_id = $1",
+            &w_id
+        )
+    .execute(&db)
+    .await?;
+    }
+
+    sqlx::query!(
+        "UPDATE workspace_settings SET mute_critical_alerts = $1 WHERE workspace_id = $2",
+        mute_alerts,
+        &w_id
+    )
+    .execute(&db)
+    .await?;
+
+    Ok(format!("Updated mute criticital alert ui settings for workspace: {}", &w_id))
+}
+
+#[cfg(not(feature = "enterprise"))]
+pub async fn mute_critical_alerts() -> Error {
+    Error::NotFound("Critical Alerts require EE".to_string())
+}

backend/windmill-common/Cargo.toml

@@ -58,6 +58,7 @@ async-stream.workspace = true
 const_format.workspace = true
 crc.workspace = true
 windmill-macros.workspace = true
+semver.workspace = true
 
 [target.'cfg(not(target_env = "msvc"))'.dependencies]
 tikv-jemalloc-ctl = { optional = true, workspace = true }

backend/windmill-common/src/error.rs

@@ -59,7 +59,7 @@ pub enum Error {
     #[error("Error: {0:#?}")]
     JsonErr(serde_json::Value),
     #[error("{0}")]
-    OpenAIError(String),
+    AiError(String),
     #[error("{0}")]
     AlreadyCompleted(String),
 }
@@ -90,7 +90,7 @@ impl IntoResponse for Error {
             Self::RequireAdmin(_) => axum::http::StatusCode::FORBIDDEN,
             Self::SqlErr(_)
             | Self::BadRequest(_)
-            | Self::OpenAIError(_)
+            | Self::AiError(_)
             | Self::QuotaExceeded(_) => axum::http::StatusCode::BAD_REQUEST,
             _ => axum::http::StatusCode::INTERNAL_SERVER_ERROR,
         };

backend/windmill-common/src/flows.rs

@@ -16,6 +16,7 @@ use rand::Rng;
 use serde::{Deserialize, Serialize, Serializer};
 
 use crate::{
+    error::Error,
     more_serde::{default_empty_string, default_id, default_null, default_true, is_default},
     scripts::{Schema, ScriptHash, ScriptLang},
 };
@@ -651,3 +652,29 @@ pub fn add_virtual_items_if_necessary(modules: &mut Vec<FlowModule>) {
         });
     }
 }
+
+pub async fn has_failure_module<'c>(flow: sqlx::types::Uuid, db: &sqlx::Pool<sqlx::Postgres>, completed: bool) -> Result<bool, Error> {
+    if completed {
+        sqlx::query_scalar!(
+            "SELECT raw_flow->'failure_module' != 'null'::jsonb
+            FROM completed_job
+            WHERE id = $1",
+            flow
+        )
+    } else {
+        sqlx::query_scalar!(
+            "SELECT raw_flow->'failure_module' != 'null'::jsonb
+            FROM queue
+            WHERE id = $1",
+            flow
+        )
+    }
+    .fetch_one(db)
+    .await
+    .map_err(|e| {
+        Error::InternalErr(format!(
+            "error during retrieval of has_failure_module: {e:#}"
+        ))
+    })
+    .map(|v| v.unwrap_or(false))
+}

backend/windmill-common/src/global_settings.rs

@@ -29,6 +29,7 @@ pub const AUTOMATE_USERNAME_CREATION_SETTING: &str = "automate_username_creation
 pub const HUB_BASE_URL_SETTING: &str = "hub_base_url";
 pub const HUB_ACCESSIBLE_URL_SETTING: &str = "hub_accessible_url";
 pub const CRITICAL_ERROR_CHANNELS_SETTING: &str = "critical_error_channels";
+pub const CRITICAL_ALERT_MUTE_UI_SETTING: &str = "critical_alert_mute_ui";
 pub const DEV_INSTANCE_SETTING: &str = "dev_instance";
 pub const JWT_SECRET_SETTING: &str = "jwt_secret";
 pub const EMAIL_DOMAIN_SETTING: &str = "email_domain";

backend/windmill-common/src/jobs.rs

@@ -108,13 +108,6 @@ pub struct QueuedJob {
     pub cache_ttl: Option<i32>,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub priority: Option<i16>,
-
-    #[serde(skip_serializing_if = "Option::is_none")]
-    #[sqlx(skip)]
-    pub self_wait_time_ms: Option<i64>,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    #[sqlx(skip)]
-    pub aggregate_wait_time_ms: Option<i64>,
 }
 
 impl QueuedJob {
@@ -198,8 +191,6 @@ impl Default for QueuedJob {
             flow_step_id: None,
             cache_ttl: None,
             priority: None,
-            self_wait_time_ms: None,
-            aggregate_wait_time_ms: None,
         }
     }
 }
@@ -253,13 +244,6 @@ pub struct CompletedJob {
     pub priority: Option<i16>,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub labels: Option<serde_json::Value>,
-
-    #[serde(skip_serializing_if = "Option::is_none")]
-    #[sqlx(skip)]
-    pub self_wait_time_ms: Option<i64>,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    #[sqlx(skip)]
-    pub aggregate_wait_time_ms: Option<i64>,
 }
 
 impl CompletedJob {

backend/windmill-common/src/lib.rs

@@ -86,6 +86,8 @@ lazy_static::lazy_static! {
     pub static ref METRICS_ENABLED: AtomicBool = AtomicBool::new(std::env::var("METRICS_PORT").is_ok() || std::env::var("METRICS_ADDR").is_ok());
     pub static ref METRICS_DEBUG_ENABLED: AtomicBool = AtomicBool::new(false);
 
+    pub static ref CRITICAL_ALERT_MUTE_UI_ENABLED: AtomicBool = AtomicBool::new(false);
+
     pub static ref BASE_URL: Arc<RwLock<String>> = Arc::new(RwLock::new("".to_string()));
     pub static ref IS_READY: std::sync::atomic::AtomicBool = std::sync::atomic::AtomicBool::new(false);
 

backend/windmill-common/src/utils.rs

@@ -21,6 +21,7 @@ use reqwest::Client;
 use serde::{Deserialize, Serialize};
 use sha2::{Digest, Sha256};
 use sqlx::{Pool, Postgres};
+use semver::Version;
 
 pub const MAX_PER_PAGE: usize = 10000;
 pub const DEFAULT_PER_PAGE: usize = 1000;
@@ -28,12 +29,22 @@ pub const DEFAULT_PER_PAGE: usize = 1000;
 pub const GIT_VERSION: &str =
     git_version!(args = ["--tag", "--always"], fallback = "unknown-version");
 
+use crate::CRITICAL_ALERT_MUTE_UI_ENABLED;
+use std::sync::atomic::Ordering;
+
+#[cfg(feature = "enterprise")]
+use crate::worker::CLOUD_HOSTED;
+
 lazy_static::lazy_static! {
     pub static ref HTTP_CLIENT: Client = reqwest::ClientBuilder::new()
         .user_agent("windmill/beta")
         .timeout(std::time::Duration::from_secs(20))
         .connect_timeout(std::time::Duration::from_secs(10))
         .build().unwrap();
+    pub static ref GIT_SEM_VERSION: Version = Version::parse(
+        // skip first `v` character.
+        GIT_VERSION.split_at(1).1
+    ).unwrap_or(Version::new(0, 1, 0));
 }
 
 #[derive(Deserialize, Clone)]
@@ -241,12 +252,40 @@ pub fn generate_lock_id(database_name: &str) -> i64 {
     0x3d32ad9e * (CRC_IEEE.checksum(database_name.as_bytes()) as i64)
 }
 
-pub async fn report_critical_error(error_message: String, _db: DB) -> () {
+pub async fn report_critical_error(
+    error_message: String,
+    _db: DB,
+    workspace_id: Option<&str>,
+    resource: Option<&str>,
+) -> () {
     tracing::error!("CRITICAL ERROR: {error_message}");
 
+    let mute_global = CRITICAL_ALERT_MUTE_UI_ENABLED.load(Ordering::Relaxed);
+    let mute_workspace = if let Some(workspace_id) = workspace_id {
+        match fetch_mute_workspace(&_db, workspace_id).await {
+            Ok(flag) => flag,
+            Err(err) => {
+                tracing::error!("Error fetching mute_workspace: {}", err);
+                false
+            }
+        }
+    } else {
+        false
+    };
+
+    // we ack_global if mute_global is true, or if mute_workspace is true
+    // but we ignore global mute setting for ack_workspace
+    let acknowledge_workspace = mute_workspace;
+    let acknowledge_global = mute_global || mute_workspace;
+
     if let Err(err) = sqlx::query!(
-        "INSERT INTO alerts (alert_type, message) VALUES ('critical_error', $1)",
-        error_message
+        "INSERT INTO alerts (alert_type, message, acknowledged, acknowledged_workspace, workspace_id, resource)
+        VALUES ('critical_error', $1, $2, $3, $4, $5)",
+        error_message,
+        acknowledge_global,
+        acknowledge_workspace,
+        workspace_id,
+        resource,
     )
     .execute(&_db)
     .await
@@ -255,27 +294,86 @@ pub async fn report_critical_error(error_message: String, _db: DB) -> () {
     }
 
     #[cfg(feature = "enterprise")]
-    send_critical_alert(error_message, &_db, CriticalAlertKind::CriticalError, None).await;
+    if *CLOUD_HOSTED && workspace_id.is_some() {
+        tracing::error!(error_message)
+    } else {
+        send_critical_alert(error_message, &_db, CriticalAlertKind::CriticalError, None).await;
+    }
 }
 
-pub async fn report_recovered_critical_error(message: String, _db: DB) -> () {
+pub async fn report_recovered_critical_error(
+    message: String,
+    _db: DB,
+    workspace_id: Option<&str>,
+    resource: Option<&str>,
+) -> () {
     tracing::info!("RECOVERED CRITICAL ERROR: {message}");
 
     if let Err(err) = sqlx::query!(
-        "INSERT INTO alerts (alert_type, message) VALUES ('recovered_critical_error', $1)",
-        message
+        "INSERT INTO alerts (alert_type, message, acknowledged, acknowledged_workspace, workspace_id, resource)
+        VALUES ('recovered_critical_error', $1, $2, $3, $4, $5)",
+        message,
+        true,
+        true,
+        workspace_id,
+        resource,
     )
     .execute(&_db)
     .await
     {
-        tracing::error!("Failed to save critical error to database: {}", err);
+        tracing::error!("Failed to save recovered critical error to database: {}", err);
     }
+
+    // acknowledge all alerts with the same resource
+    if let Some(resource) = resource {
+        if let Err(err) = sqlx::query!(
+            "UPDATE alerts SET acknowledged = true, acknowledged_workspace = true WHERE resource = $1 AND alert_type = 'critical_error'",
+            resource,
+        )
+        .execute(&_db)
+        .await
+        {
+            tracing::error!("Failed to acknowledge critical error alerts for resource {}: {}", resource, err);
+        }
+    }
+
     #[cfg(feature = "enterprise")]
-    send_critical_alert(
-        message,
-        &_db,
-        CriticalAlertKind::RecoveredCriticalError,
-        None,
-    )
-    .await;
+    if *CLOUD_HOSTED && workspace_id.is_some() {
+        tracing::error!(message);
+    } else {
+        send_critical_alert(
+            message,
+            &_db,
+            CriticalAlertKind::RecoveredCriticalError,
+            None,
+        )
+        .await;
+    }
+}
+
+pub async fn fetch_mute_workspace(_db: &DB, workspace_id: &str) -> Result<bool> {
+    match sqlx::query!(
+        "SELECT mute_critical_alerts FROM workspace_settings WHERE workspace_id = $1",
+        workspace_id
+    )
+    .fetch_optional(_db)
+    .await
+    {
+        Ok(Some(record)) => Ok(record.mute_critical_alerts.unwrap_or(false)),
+        Ok(None) => {
+            tracing::warn!(
+                "Workspace ID {} not found in workspace_settings table",
+                workspace_id
+            );
+            Ok(false)
+        }
+        Err(err) => {
+            tracing::error!(
+                "Error querying workspace_settings for workspace_id {}: {}",
+                workspace_id,
+                err
+            );
+            Err(Error::SqlErr(err))
+        }
+    }
 }

backend/windmill-common/src/worker.rs

@@ -1,6 +1,7 @@
 use const_format::concatcp;
 use itertools::Itertools;
 use regex::Regex;
+use semver::Version;
 use serde::{Deserialize, Serialize};
 use serde_json::value::RawValue;
 use std::{
@@ -88,6 +89,7 @@ lazy_static::lazy_static! {
     .and_then(|x| x.parse::<bool>().ok())
     .unwrap_or(false);
 
+    pub static ref MIN_VERSION: Arc<RwLock<Version>> = Arc::new(RwLock::new(Version::new(0, 0, 0)));
 }
 
 pub async fn make_suspended_pull_query(wc: &WorkerConfig) {
@@ -308,6 +310,8 @@ fn parse_file<T: FromStr>(path: &str) -> Option<T> {
 pub struct PythonAnnotations {
     pub no_cache: bool,
     pub no_uv: bool,
+    pub no_uv_install: bool,
+    pub no_uv_compile: bool,
 }
 
 #[annotations("//")]
@@ -548,6 +552,31 @@ pub fn get_windmill_memory_usage() -> Option<i64> {
     }
 }
 
+pub async fn update_min_version<'c, E: sqlx::Executor<'c, Database = sqlx::Postgres>>(executor: E) -> bool {
+    use crate::utils::{GIT_VERSION, GIT_SEM_VERSION};
+
+    // fetch all pings with a different version than self from the last 5 minutes.
+    let pings = sqlx::query_scalar!(
+        "SELECT wm_version FROM worker_ping WHERE wm_version != $1 AND ping_at > now() - interval '5 minutes'",
+        GIT_VERSION
+    ).fetch_all(executor).await.unwrap_or_default();
+
+    let cur_version = GIT_SEM_VERSION.clone();
+    let min_version = pings
+        .iter()
+        .filter(|x| !x.is_empty())
+        .filter_map(|x| semver::Version::parse(x.split_at(1).1).ok())
+        .min()
+        .unwrap_or_else(|| cur_version.clone());
+
+    if min_version != cur_version {
+        tracing::info!("Minimal worker version: {min_version}");
+    }
+
+    *MIN_VERSION.write().await = min_version.clone();
+    min_version >= cur_version
+}
+
 pub async fn update_ping(worker_instance: &str, worker_name: &str, ip: &str, db: &DB) {
     let (tags, dw) = {
         let wc = WORKER_CONFIG.read().await.clone();

backend/windmill-queue/src/jobs.rs

@@ -71,6 +71,9 @@ use windmill_common::BASE_URL;
 #[cfg(feature = "cloud")]
 use windmill_common::users::SUPERADMIN_SYNC_EMAIL;
 
+#[cfg(feature = "enterprise")]
+use windmill_common::flows::has_failure_module;
+
 #[cfg(feature = "enterprise")]
 use windmill_common::worker::CLOUD_HOSTED;
 
@@ -482,13 +485,6 @@ where
     }
 }
 
-#[cfg(feature = "enterprise")]
-#[derive(Deserialize)]
-struct RawFlowFailureModule {
-    #[cfg(feature = "enterprise")]
-    failure_module: Option<Box<RawValue>>,
-}
-
 #[instrument(level = "trace", skip_all)]
 pub async fn add_completed_job_error<R: rsmq_async::RsmqConnection + Clone + Send>(
     db: &Pool<Postgres>,
@@ -928,6 +924,8 @@ pub async fn add_completed_job<
                         .unwrap_or("".to_string()),
                 ),
                 db.clone(),
+                Some(&w_id),
+                None,
             )
             .await;
         } else if queued_job.email == SCHEDULE_ERROR_HANDLER_USER_EMAIL {
@@ -952,15 +950,7 @@ pub async fn add_completed_job<
         } else if !skip_downstream_error_handlers
             && (matches!(queued_job.job_kind, JobKind::Script)
                 || matches!(queued_job.job_kind, JobKind::Flow)
-                    && queued_job
-                        .raw_flow
-                        .as_ref()
-                        .and_then(|v| {
-                            serde_json::from_str::<RawFlowFailureModule>((**v).get())
-                                .ok()
-                                .and_then(|v| v.failure_module)
-                        })
-                        .is_none())
+                    && !has_failure_module(job_id, db, true).await.unwrap_or(false))
             && queued_job.parent_job.is_none()
         {
             let result = serde_json::from_str(
@@ -1004,7 +994,7 @@ pub async fn add_completed_job<
                             "Could not push workspace error handler for failed job ({base_url}/run/{}?workspace={w_id}): {}",
                             queued_job.id,
                             err
-                        ), db.clone())
+                        ), db.clone(), Some(&w_id), None)
                         .await;
                     }
                 }
@@ -1187,10 +1177,10 @@ pub async fn report_error_to_workspace_handler_or_critical_side_channel<
                 queued_job.id,
                 err
             );
-            report_critical_error(error_message, db.clone()).await;
+            report_critical_error(error_message, db.clone(), Some(&w_id), None).await;
         }
     } else {
-        report_critical_error(error_message, db.clone()).await;
+        report_critical_error(error_message, db.clone(), Some(&w_id), None).await;
     }
 }
 
@@ -3682,15 +3672,15 @@ pub async fn push<'c, 'd, R: rsmq_async::RsmqConnection + Send + 'c>(
         }
         JobPayload::DeploymentCallback { path } => (
             None,
-            Some(path),
+            Some(path.clone()),
             None,
             JobKind::DeploymentCallback,
             None,
             None,
             None,
-            None,
-            None,
-            None,
+            Some(format!("{workspace_id}:git_sync")),
+            Some(1),
+            Some(0),
             None,
             None,
             None,

backend/windmill-worker/nsjail/download.py.pip.config.proto

@@ -0,0 +1,93 @@
+name: "python download pip"
+
+mode: ONCE
+hostname: "python"
+log_level: ERROR
+time_limit: 900
+
+rlimit_as: 2048
+rlimit_cpu: 1000
+rlimit_fsize: 1024
+rlimit_nofile: 64
+
+envar: "HOME=/user"
+envar: "LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH"
+
+cwd: "/tmp"
+
+clone_newnet: false
+clone_newuser: {CLONE_NEWUSER}
+
+keep_caps: true
+keep_env: true
+
+mount {
+    src: "/bin"
+    dst: "/bin"
+	is_bind: true
+}
+
+mount {
+    src: "/lib"
+    dst: "/lib"
+	is_bind: true
+}
+
+mount {
+    src: "/lib64"
+    dst: "/lib64"
+	is_bind: true
+    mandatory: false
+}
+
+mount {
+    src: "/usr"
+    dst: "/usr"
+	is_bind: true
+}
+
+mount {
+    src: "/etc"
+    dst: "/etc"
+	is_bind: true
+}
+
+mount {
+	src: "/dev/null"
+	dst: "/dev/null"
+	is_bind: true
+	rw: true
+}
+
+mount {
+	dst: "/tmp"
+	fstype: "tmpfs"
+	rw: true
+    options: "size=500000000"
+}
+
+
+mount {
+    src: "{WORKER_DIR}/download_deps.py.pip.sh"
+    dst: "/download_deps.sh"
+    is_bind: true
+}
+
+mount {
+    src: "{CACHE_DIR}"
+    dst: "{CACHE_DIR}"
+    is_bind: true
+    rw: true
+}
+
+mount {
+    src: "/dev/urandom"
+    dst: "/dev/urandom"
+    is_bind: true
+}
+
+exec_bin {
+ 	path: "/bin/sh"
+ 	arg: "/download_deps.sh"
+}
+

backend/windmill-worker/nsjail/download_deps.py.pip.sh

@@ -0,0 +1,24 @@
+#/bin/sh
+
+INDEX_URL_ARG=$([ -z "$INDEX_URL" ] && echo ""|| echo "--index-url $INDEX_URL" )
+EXTRA_INDEX_URL_ARG=$([ -z "$EXTRA_INDEX_URL" ] && echo ""|| echo "--extra-index-url $EXTRA_INDEX_URL" )
+TRUSTED_HOST_ARG=$([ -z "$TRUSTED_HOST" ] &&  echo "" || echo "--trusted-host $TRUSTED_HOST")
+
+if [ ! -z "$INDEX_URL" ]
+then
+      echo "\$INDEX_URL is set to $INDEX_URL"
+fi
+
+if [ ! -z "$EXTRA_INDEX_URL" ]
+then
+      echo "\$EXTRA_INDEX_URL is set to $EXTRA_INDEX_URL"
+fi
+
+if [ ! -z "$TRUSTED_HOST" ]
+then
+      echo "\$TRUSTED_HOST is set to $TRUSTED_HOST"
+fi
+
+CMD="/usr/local/bin/python3 -m pip install -v \"$REQ\" -I -t \"$TARGET\" --no-cache --no-color --no-deps --isolated --no-warn-conflicts --disable-pip-version-check $INDEX_URL_ARG $EXTRA_INDEX_URL_ARG $TRUSTED_HOST_ARG"
+echo $CMD
+eval $CMD

backend/windmill-worker/nsjail/download_deps.py.sh

@@ -19,6 +19,18 @@ then
       echo "\$TRUSTED_HOST is set to $TRUSTED_HOST"
 fi
 
-CMD="/usr/local/bin/python3 -m pip install -v \"$REQ\" -I -t \"$TARGET\" --no-cache --no-color --no-deps --isolated --no-warn-conflicts --disable-pip-version-check $INDEX_URL_ARG $EXTRA_INDEX_URL_ARG $TRUSTED_HOST_ARG"
+CMD="/usr/local/bin/uv pip install 
+\"$REQ\" 
+--target \"$TARGET\" 
+--no-cache 
+--no-config
+--no-color 
+--no-deps 
+--link-mode=copy
+$INDEX_URL_ARG $EXTRA_INDEX_URL_ARG $TRUSTED_HOST_ARG
+--index-strategy unsafe-best-match
+--system
+"
+
 echo $CMD
 eval $CMD

backend/windmill-worker/src/ansible_executor.rs

@@ -116,6 +116,8 @@ async fn handle_ansible_python_deps(
             job_dir,
             worker_dir,
             &mut Some(occupancy_metrics),
+            true,
+            true,
         )
         .await?;
         additional_python_paths.append(&mut venv_path);

backend/windmill-worker/src/bash_executor.rs

@@ -89,8 +89,18 @@ cat bp | tail -1 >> ./result2.out &
 # Run main.sh in the same process group
 {bash} ./main.sh "$@" 2>&1 | tee bp &
 
-# Wait for all background processes to finish
-wait
+pid=$!
+
+# Wait for main.sh to finish and capture its exit status
+wait $pid
+exit_status=$?
+
+# Clean up the named pipe and background processes
+rm -f bp
+
+
+# Exit with the captured status
+exit $exit_status
 "#,
         bash = BIN_BASH.as_str(),
     );

backend/windmill-worker/src/common.rs

@@ -41,7 +41,7 @@ use tokio::{io::AsyncWriteExt, process::Child, time::Instant};
 
 use crate::{
     AuthedClient, AuthedClientBackgroundTask, JOB_DEFAULT_TIMEOUT, MAX_RESULT_SIZE,
-    MAX_TIMEOUT_DURATION,
+    MAX_TIMEOUT_DURATION, PATH_ENV,
 };
 
 pub async fn build_args_map<'a>(
@@ -860,11 +860,17 @@ pub async fn save_in_cache(
 }
 
 fn tentatively_improve_error(err: Error, executable: &str) -> Error {
+    #[cfg(unix)]
+    let err_msg = "No such file or directory (os error 2)";
+
+    #[cfg(windows)]
+    let err_msg = "program not found";
+
     if err
         .to_string()
-        .contains("No such file or directory (os error 2)")
+        .contains(&err_msg)
     {
-        return Error::InternalErr(format!("Executable {executable} not found on worker"));
+        return Error::InternalErr(format!("Executable {executable} not found on worker. PATH: {}", *PATH_ENV));
     }
     return err;
 }

backend/windmill-worker/src/global_cache.rs

@@ -20,13 +20,21 @@ use std::sync::Arc;
 pub async fn build_tar_and_push(
     s3_client: Arc<dyn ObjectStore>,
     folder: String,
+    no_uv: bool,
 ) -> error::Result<()> {
     use object_store::path::Path;
 
+    use crate::PY311_CACHE_DIR;
+
     tracing::info!("Started building and pushing piptar {folder}");
     let start = Instant::now();
     let folder_name = folder.split("/").last().unwrap();
-    let tar_path = format!("{PIP_CACHE_DIR}/{folder_name}_tar.tar",);
+    let prefix = if no_uv {
+        PIP_CACHE_DIR
+    } else {
+        PY311_CACHE_DIR
+    };
+    let tar_path = format!("{prefix}/{folder_name}_tar.tar",);
 
     let tar_file = std::fs::File::create(&tar_path)?;
     let mut tar = tar::Builder::new(tar_file);
@@ -46,7 +54,10 @@ pub async fn build_tar_and_push(
     // })?;
     if let Err(e) = s3_client
         .put(
-            &Path::from(format!("/tar/pip/{folder_name}.tar")),
+            &Path::from(format!(
+                "/tar/{}/{folder_name}.tar",
+                if no_uv { "pip" } else { "python_311" }
+            )),
             std::fs::read(&tar_path)?.into(),
         )
         .await
@@ -71,7 +82,11 @@ pub async fn build_tar_and_push(
 }
 
 #[cfg(all(feature = "enterprise", feature = "parquet"))]
-pub async fn pull_from_tar(client: Arc<dyn ObjectStore>, folder: String) -> error::Result<()> {
+pub async fn pull_from_tar(
+    client: Arc<dyn ObjectStore>,
+    folder: String,
+    no_uv: bool,
+) -> error::Result<()> {
     use windmill_common::s3_helpers::attempt_fetch_bytes;
 
     let folder_name = folder.split("/").last().unwrap();
@@ -79,7 +94,10 @@ pub async fn pull_from_tar(client: Arc<dyn ObjectStore>, folder: String) -> erro
     tracing::info!("Attempting to pull piptar {folder_name} from bucket");
 
     let start = Instant::now();
-    let tar_path = format!("tar/pip/{folder_name}.tar");
+    let tar_path = format!(
+        "tar/{}/{folder_name}.tar",
+        if no_uv { "pip" } else { "python_311" }
+    );
     let bytes = attempt_fetch_bytes(client, &tar_path).await?;
 
     // tracing::info!("B: {target} {folder}");

backend/windmill-worker/src/handle_child.rs

@@ -48,6 +48,7 @@ use futures::{
 
 use crate::common::{resolve_job_timeout, OccupancyMetrics};
 use crate::job_logger::{append_job_logs, append_with_limit, LARGE_LOG_THRESHOLD_SIZE};
+use crate::job_logger_ee::process_streaming_log_lines;
 use crate::{MAX_RESULT_SIZE, MAX_WAIT_FOR_SIGINT, MAX_WAIT_FOR_SIGTERM};
 
 lazy_static::lazy_static! {
@@ -434,6 +435,8 @@ pub async fn handle_child(
     }
 }
 
+
+
 async fn get_mem_peak(pid: Option<u32>, nsjail: bool) -> i32 {
     if pid.is_none() {
         return -1;
@@ -689,19 +692,24 @@ fn child_joined_output_stream(
 
     let stdout = BufReader::new(stdout).lines();
     let stderr = BufReader::new(stderr).lines();
-    stream::select(lines_to_stream(stderr), lines_to_stream(stdout))
+    stream::select(lines_to_stream(stderr, true), lines_to_stream(stdout, false))
 }
 
 pub fn lines_to_stream<R: tokio::io::AsyncBufRead + Unpin>(
     mut lines: tokio::io::Lines<R>,
+    stderr: bool,
 ) -> impl futures::Stream<Item = io::Result<String>> {
     stream::poll_fn(move |cx| {
         std::pin::Pin::new(&mut lines)
             .poll_next_line(cx)
-            .map(|result| result.transpose())
+            .map(|result| {
+                process_streaming_log_lines(result, stderr)
+             })
     })
 }
 
+
+
 pub fn process_status(status: ExitStatus) -> error::Result<()> {
     if status.success() {
         Ok(())

backend/windmill-worker/src/job_logger.rs

@@ -1,15 +1,6 @@
-use itertools::Itertools;
-use swc_ecma_parser::lexer::util::CharExt;
-
-#[cfg(all(feature = "enterprise", feature = "parquet"))]
-use object_store::path::Path;
 use regex::Regex;
 
-#[cfg(all(feature = "enterprise", feature = "parquet"))]
-use windmill_common::s3_helpers::OBJECT_STORE_CACHE_SETTINGS;
-
-use windmill_common::error::{self};
-use windmill_common::worker::{CLOUD_HOSTED, TMP_DIR};
+use windmill_common::worker::CLOUD_HOSTED;
 
 use windmill_queue::append_logs;
 
@@ -19,6 +10,12 @@ use std::sync::Arc;
 use uuid::Uuid;
 use windmill_common::DB;
 
+use crate::job_logger_ee::default_disk_log_storage;
+
+#[cfg(all(feature = "enterprise", feature = "parquet"))]
+use crate::job_logger_ee::s3_storage;
+
+
 pub enum CompactLogs {
     #[cfg(not(all(feature = "enterprise", feature = "parquet")))]
     NotEE,
@@ -28,150 +25,7 @@ pub enum CompactLogs {
     S3,
 }
 
-async fn compact_logs(
-    job_id: Uuid,
-    w_id: &str,
-    db: &DB,
-    nlogs: String,
-    total_size: Arc<AtomicU32>,
-    compact_kind: CompactLogs,
-    _worker_name: &str,
-) -> error::Result<(String, String)> {
-    let mut prev_logs = sqlx::query_scalar!(
-        "SELECT logs FROM job_logs WHERE job_id = $1 AND workspace_id = $2",
-        job_id,
-        w_id
-    )
-    .fetch_optional(db)
-    .await?
-    .flatten()
-    .unwrap_or_default();
-    let size = prev_logs.char_indices().count() as i32;
-    let nlogs_len = nlogs.char_indices().count();
-    let to_keep_in_db = usize::max(
-        usize::min(nlogs_len, 3000),
-        nlogs_len % LARGE_LOG_THRESHOLD_SIZE,
-    );
-    let extra_split = to_keep_in_db < nlogs_len;
-    let stored_in_storage_len = if extra_split {
-        nlogs_len - to_keep_in_db
-    } else {
-        0
-    };
-    let extra_to_newline = nlogs
-        .chars()
-        .skip(stored_in_storage_len)
-        .find_position(|x| x.is_line_break())
-        .map(|(i, _)| i)
-        .unwrap_or(to_keep_in_db);
-    let stored_in_storage_to_newline = stored_in_storage_len + extra_to_newline;
 
-    let (append_to_storage, stored_in_db) = if extra_split {
-        if stored_in_storage_to_newline == nlogs.len() {
-            (nlogs.as_ref(), "".to_string())
-        } else {
-            let split_idx = nlogs
-                .char_indices()
-                .nth(stored_in_storage_to_newline)
-                .map(|(i, _)| i)
-                .unwrap_or(0);
-            let (append_to_storage, stored_in_db) = nlogs.split_at(split_idx);
-            // tracing::error!("{append_to_storage} ||||| {stored_in_db}");
-            // tracing::error!(
-            //     "{:?} {:?} {} {}",
-            //     excess_prev_logs.lines().last(),
-            //     current_logs.lines().next(),
-            //     split_idx,
-            //     excess_size_modulo
-            // );
-            (append_to_storage, stored_in_db.to_string())
-        }
-    } else {
-        // tracing::error!("{:?}", nlogs.lines().last());
-        ("", nlogs.to_string())
-    };
-
-    let new_size_with_excess = size + stored_in_storage_to_newline as i32;
-
-    let new_size = total_size.fetch_add(
-        new_size_with_excess as u32,
-        std::sync::atomic::Ordering::SeqCst,
-    ) + new_size_with_excess as u32;
-
-    let path = format!(
-        "logs/{job_id}/{}_{new_size}.txt",
-        chrono::Utc::now().timestamp_millis()
-    );
-
-    let mut new_current_logs = match compact_kind {
-        CompactLogs::NoS3 => format!("\n[windmill] No object storage set in instance settings. Previous logs have been saved to disk at {path}"),
-        CompactLogs::S3 => format!("\n[windmill] Previous logs have been saved to object storage at {path}"),
-        #[cfg(not(all(feature = "enterprise", feature = "parquet")))]
-        CompactLogs::NotEE => format!("\n[windmill] Previous logs have been saved to disk at {path}"),
-    };
-    new_current_logs.push_str(&stored_in_db);
-
-    sqlx::query!(
-        "UPDATE job_logs SET logs = $1, log_offset = $2, 
-        log_file_index = array_append(coalesce(log_file_index, array[]::text[]), $3) 
-        WHERE workspace_id = $4 AND job_id = $5",
-        new_current_logs,
-        new_size as i32,
-        path,
-        w_id,
-        job_id
-    )
-    .execute(db)
-    .await?;
-    prev_logs.push_str(&append_to_storage);
-
-    return Ok((prev_logs, path));
-}
-
-async fn default_disk_log_storage(
-    job_id: Uuid,
-    w_id: &str,
-    db: &DB,
-    nlogs: String,
-    total_size: Arc<AtomicU32>,
-    compact_kind: CompactLogs,
-    worker_name: &str,
-) {
-    match compact_logs(
-        job_id,
-        &w_id,
-        &db,
-        nlogs,
-        total_size,
-        compact_kind,
-        worker_name,
-    )
-    .await
-    {
-        Err(e) => tracing::error!("Could not compact logs for job {job_id}: {e:?}",),
-        Ok((prev_logs, path)) => {
-            let path = format!("{}/{}", TMP_DIR, path);
-            let splitted = &path.split("/").collect_vec();
-            tokio::fs::create_dir_all(splitted.into_iter().take(splitted.len() - 1).join("/"))
-                .await
-                .map_err(|e| {
-                    tracing::error!("Could not create logs directory: {e:?}",);
-                    e
-                })
-                .ok();
-            let created = tokio::fs::File::create(&path).await;
-            if let Err(e) = created {
-                tracing::error!("Could not create logs file {path}: {e:?}",);
-                return;
-            }
-            if let Err(e) = tokio::fs::write(&path, prev_logs).await {
-                tracing::error!("Could not write to logs file {path}: {e:?}");
-            } else {
-                tracing::info!("Logs length of {job_id} has exceeded a threshold. Previous logs have been saved to disk at {path}");
-            }
-        }
-    }
-}
 
 pub(crate) async fn append_job_logs(
     job_id: Uuid,
@@ -184,43 +38,7 @@ pub(crate) async fn append_job_logs(
 ) -> () {
     if must_compact_logs {
         #[cfg(all(feature = "enterprise", feature = "parquet"))]
-        if let Some(os) = OBJECT_STORE_CACHE_SETTINGS.read().await.clone() {
-            match compact_logs(
-                job_id,
-                &w_id,
-                &db,
-                logs,
-                total_size,
-                CompactLogs::S3,
-                &worker_name,
-            )
-            .await
-            {
-                Err(e) => tracing::error!("Could not compact logs for job {job_id}: {e:?}",),
-                Ok((prev_logs, path)) => {
-                    tracing::info!("Logs length of {job_id} has exceeded a threshold. Previous logs have been saved to object storage at {path}");
-                    let path2 = path.clone();
-                    if let Err(e) = os
-                        .put(&Path::from(path), prev_logs.to_string().into_bytes().into())
-                        .await
-                    {
-                        tracing::error!("Could not save logs to s3: {e:?}");
-                    }
-                    tracing::info!("Logs of {job_id} saved to object storage at {path2}");
-                }
-            }
-        } else {
-            default_disk_log_storage(
-                job_id,
-                &w_id,
-                &db,
-                logs,
-                total_size,
-                CompactLogs::NoS3,
-                &worker_name,
-            )
-            .await;
-        }
+        s3_storage(job_id, &w_id, &db, logs, total_size, &worker_name).await;
 
         #[cfg(not(all(feature = "enterprise", feature = "parquet")))]
         {
@@ -252,6 +70,7 @@ pub fn append_with_limit(dst: &mut String, src: &str, limit: &mut usize) {
     if *NO_LOGS_AT_ALL {
         return;
     }
+    
     let src_str;
     let src = {
         src_str = RE_00.replace_all(src, "");

backend/windmill-worker/src/job_logger_ee.rs

@@ -0,0 +1,30 @@
+use std::io;
+use std::sync::atomic::AtomicU32;
+use std::sync::Arc;
+
+use uuid::Uuid;
+use windmill_common::DB;
+
+use crate::job_logger::CompactLogs;
+
+#[cfg(all(feature = "enterprise", feature = "parquet"))]
+pub(crate) async fn s3_storage(_job_id: Uuid, _w_id: &String, _db: &sqlx::Pool<sqlx::Postgres>, _logs: &String, _total_size: &Arc<AtomicU32>, _worker_name: &String) {
+    tracing::info!("Logs length of {job_id} has exceeded a threshold. Implementation to store excess on s3 in not OSS");
+}
+
+pub(crate) async fn default_disk_log_storage(
+    job_id: Uuid,
+    _w_id: &str,
+    _db: &DB,
+    _nlogs: String,
+    _total_size: Arc<AtomicU32>,
+    _compact_kind: CompactLogs,
+    _worker_name: &str,
+) {
+    tracing::info!("Logs length of {job_id} has exceeded a threshold. Implementation to store excess on disk in not OSS");
+}
+
+
+pub(crate) fn process_streaming_log_lines(r: Result<Option<String>, io::Error>, _stderr: bool) -> Option<Result<String, io::Error>> {
+    r.transpose()
+}

backend/windmill-worker/src/lib.rs

@@ -29,7 +29,7 @@ mod rust_executor;
 mod worker;
 mod worker_flow;
 mod worker_lockfiles;
-
+mod job_logger_ee;
 pub use worker::*;
 
 pub use result_processor::handle_job_error;

backend/windmill-worker/src/python_executor.rs

@@ -1,4 +1,4 @@
-use std::{collections::HashMap, process::Stdio};
+use std::{collections::HashMap, fs, process::Stdio};
 
 use itertools::Itertools;
 use regex::Regex;
@@ -42,6 +42,10 @@ lazy_static::lazy_static! {
     static ref USE_PIP_COMPILE: bool = std::env::var("USE_PIP_COMPILE")
         .ok().map(|flag| flag == "true").unwrap_or(false);
 
+    /// Use pip install
+    static ref USE_PIP_INSTALL: bool = std::env::var("USE_PIP_INSTALL")
+        .ok().map(|flag| flag == "true").unwrap_or(false);
+
 
     static ref RELATIVE_IMPORT_REGEX: Regex = Regex::new(r#"(import|from)\s(((u|f)\.)|\.)"#).unwrap();
 
@@ -50,6 +54,8 @@ lazy_static::lazy_static! {
 }
 
 const NSJAIL_CONFIG_DOWNLOAD_PY_CONTENT: &str = include_str!("../nsjail/download.py.config.proto");
+const NSJAIL_CONFIG_DOWNLOAD_PY_CONTENT_FALLBACK: &str =
+    include_str!("../nsjail/download.py.pip.config.proto");
 const NSJAIL_CONFIG_RUN_PYTHON3_CONTENT: &str = include_str!("../nsjail/run.python3.config.proto");
 const RELATIVE_PYTHON_LOADER: &str = include_str!("../loader.py");
 
@@ -66,8 +72,8 @@ use crate::{
     },
     handle_child::handle_child,
     AuthedClientBackgroundTask, DISABLE_NSJAIL, DISABLE_NUSER, HOME_ENV, LOCK_CACHE_DIR,
-    NSJAIL_PATH, PATH_ENV, PIP_CACHE_DIR, PIP_EXTRA_INDEX_URL, PIP_INDEX_URL, PROXY_ENVS, TZ_ENV,
-    UV_CACHE_DIR,
+    NSJAIL_PATH, PATH_ENV, PIP_CACHE_DIR, PIP_EXTRA_INDEX_URL, PIP_INDEX_URL, PROXY_ENVS,
+    PY311_CACHE_DIR, TZ_ENV, UV_CACHE_DIR,
 };
 
 #[cfg(windows)]
@@ -199,8 +205,11 @@ pub async fn uv_pip_compile(
             .clone()
             .map(handle_ephemeral_token);
         if let Some(url) = pip_extra_index_url.as_ref() {
-            args.extend(["--extra-index-url", url, "--no-emit-index-url"]);
-            pip_args.push(format!("--extra-index-url {}", url));
+            url.split(",").for_each(|url| {
+                args.extend(["--extra-index-url", url]);
+                pip_args.push(format!("--extra-index-url {}", url));
+            });
+            args.push("--no-emit-index-url");
         }
         let pip_index_url = PIP_INDEX_URL
             .read()
@@ -279,7 +288,9 @@ pub async fn uv_pip_compile(
             .clone()
             .map(handle_ephemeral_token);
         if let Some(url) = pip_extra_index_url.as_ref() {
-            args.extend(["--extra-index-url", url]);
+            url.split(",").for_each(|url| {
+                args.extend(["--extra-index-url", url]);
+            });
         }
         let pip_index_url = PIP_INDEX_URL
             .read()
@@ -309,7 +320,7 @@ pub async fn uv_pip_compile(
             .args(args)
             .stdout(Stdio::piped())
             .stderr(Stdio::piped());
-        let child_process = start_child_process(child_cmd, "/usr/local/bin/uv").await?;
+        let child_process = start_child_process(child_cmd, uv_cmd).await?;
         append_logs(&job_id, &w_id, logs, db).await;
         handle_child(
             job_id,
@@ -896,10 +907,10 @@ async fn handle_python_deps(
         .unwrap_or_else(|| vec![])
         .clone();
 
+    let annotations = windmill_common::worker::PythonAnnotations::parse(inner_content);
     let requirements = match requirements_o {
         Some(r) => r,
         None => {
-            let annotation = windmill_common::worker::PythonAnnotations::parse(inner_content);
             let mut already_visited = vec![];
 
             let requirements = windmill_parser_py_imports::parse_python_imports(
@@ -924,8 +935,8 @@ async fn handle_python_deps(
                     worker_name,
                     w_id,
                     occupancy_metrics,
-                    annotation.no_uv,
-                    annotation.no_cache,
+                    annotations.no_uv || annotations.no_uv_compile,
+                    annotations.no_cache,
                 )
                 .await
                 .map_err(|e| {
@@ -950,6 +961,8 @@ async fn handle_python_deps(
             job_dir,
             worker_dir,
             occupancy_metrics,
+            annotations.no_uv || annotations.no_uv_install,
+            false,
         )
         .await?;
         additional_python_paths.append(&mut venv_path);
@@ -961,6 +974,7 @@ lazy_static::lazy_static! {
     static ref PIP_SECRET_VARIABLE: Regex = Regex::new(r"\$\{PIP_SECRET:([^\s\}]+)\}").unwrap();
 }
 
+/// pip install, include cached or pull from S3
 pub async fn handle_python_reqs(
     requirements: Vec<&str>,
     job_id: &Uuid,
@@ -972,12 +986,22 @@ pub async fn handle_python_reqs(
     job_dir: &str,
     worker_dir: &str,
     occupancy_metrics: &mut Option<&mut OccupancyMetrics>,
+    // TODO: Remove (Deprecated)
+    mut no_uv_install: bool,
+    is_ansible: bool,
 ) -> error::Result<Vec<String>> {
     let mut req_paths: Vec<String> = vec![];
     let mut vars = vec![("PATH", PATH_ENV.as_str())];
     let pip_extra_index_url;
     let pip_index_url;
 
+    no_uv_install |= *USE_PIP_INSTALL;
+
+    if no_uv_install && !is_ansible {
+        append_logs(&job_id, w_id, "\nFallback to pip (Deprecated!)\n", db).await;
+        tracing::warn!("Fallback to pip");
+    }
+
     if !*DISABLE_NSJAIL {
         pip_extra_index_url = PIP_EXTRA_INDEX_URL
             .read()
@@ -1008,10 +1032,21 @@ pub async fn handle_python_reqs(
         let _ = write_file(
             job_dir,
             "download.config.proto",
-            &NSJAIL_CONFIG_DOWNLOAD_PY_CONTENT
-                .replace("{WORKER_DIR}", &worker_dir)
-                .replace("{CACHE_DIR}", PIP_CACHE_DIR)
-                .replace("{CLONE_NEWUSER}", &(!*DISABLE_NUSER).to_string()),
+            &(if no_uv_install {
+                NSJAIL_CONFIG_DOWNLOAD_PY_CONTENT_FALLBACK
+            } else {
+                NSJAIL_CONFIG_DOWNLOAD_PY_CONTENT
+            })
+            .replace("{WORKER_DIR}", &worker_dir)
+            .replace(
+                "{CACHE_DIR}",
+                if no_uv_install {
+                    PIP_CACHE_DIR
+                } else {
+                    PY311_CACHE_DIR
+                },
+            )
+            .replace("{CLONE_NEWUSER}", &(!*DISABLE_NUSER).to_string()),
         )?;
     };
 
@@ -1021,8 +1056,14 @@ pub async fn handle_python_reqs(
         if req.starts_with('#') {
             continue;
         }
+        let py_prefix = if no_uv_install {
+            PIP_CACHE_DIR
+        } else {
+            PY311_CACHE_DIR
+        };
+
         let venv_p = format!(
-            "{PIP_CACHE_DIR}/{}",
+            "{py_prefix}/{}",
             req.replace(' ', "").replace('/', "").replace(':', "")
         );
         if metadata(&venv_p).await.is_ok() {
@@ -1068,7 +1109,10 @@ pub async fn handle_python_reqs(
                 .map(|(req, venv_p)| {
                     let os = os.clone();
                     async move {
-                        if pull_from_tar(os, venv_p.clone()).await.is_ok() {
+                        if pull_from_tar(os, venv_p.clone(), no_uv_install)
+                            .await
+                            .is_ok()
+                        {
                             PullFromTar::Pulled(venv_p.to_string())
                         } else {
                             PullFromTar::NotPulled(req.to_string(), venv_p.to_string())
@@ -1109,7 +1153,11 @@ pub async fn handle_python_reqs(
 
     for (req, venv_p) in req_with_penv {
         let mut logs1 = String::new();
-        logs1.push_str("\n\n--- PIP INSTALL ---\n");
+        if no_uv_install {
+            logs1.push_str("\n\n--- PIP INSTALL ---\n");
+        } else {
+            logs1.push_str("\n\n--- UV PIP INSTALL ---\n");
+        }
         logs1.push_str(&format!("\n{req} is being installed for the first time.\n It will be cached for all ulterior uses."));
         append_logs(&job_id, w_id, logs1, db).await;
 
@@ -1145,21 +1193,47 @@ pub async fn handle_python_reqs(
             #[cfg(windows)]
             let req = format!("{}", req);
 
-            let mut command_args = vec![
-                PYTHON_PATH.as_str(),
-                "-m",
-                "pip",
-                "install",
-                &req,
-                "-I",
-                "--no-deps",
-                "--no-color",
-                "--isolated",
-                "--no-warn-conflicts",
-                "--disable-pip-version-check",
-                "-t",
-                venv_p.as_str(),
-            ];
+            let mut command_args = if no_uv_install {
+                vec![
+                    PYTHON_PATH.as_str(),
+                    "-m",
+                    "pip",
+                    "install",
+                    &req,
+                    "-I",
+                    "--no-deps",
+                    "--no-color",
+                    "--isolated",
+                    "--no-warn-conflicts",
+                    "--disable-pip-version-check",
+                    "-t",
+                    venv_p.as_str(),
+                ]
+            } else {
+                vec![
+                    UV_PATH.as_str(),
+                    "pip",
+                    "install",
+                    &req,
+                    "--no-deps",
+                    "--no-color",
+                    // "-p",
+                    // "3.11",
+                    // Prevent uv from discovering configuration files.
+                    "--no-config",
+                    "--link-mode=copy",
+                    // TODO: Doublecheck it
+                    "--system",
+                    // Prefer main index over extra
+                    // https://docs.astral.sh/uv/pip/compatibility/#packages-that-exist-on-multiple-indexes
+                    // TODO: Use env variable that can be toggled from UI
+                    "--index-strategy",
+                    "unsafe-best-match",
+                    "--target",
+                    venv_p.as_str(),
+                    "--no-cache",
+                ]
+            };
             let pip_extra_index_url = PIP_EXTRA_INDEX_URL
                 .read()
                 .await
@@ -1167,7 +1241,9 @@ pub async fn handle_python_reqs(
                 .map(handle_ephemeral_token);
 
             if let Some(url) = pip_extra_index_url.as_ref() {
-                command_args.extend(["--extra-index-url", url]);
+                url.split(",").for_each(|url| {
+                    command_args.extend(["--extra-index-url", url]);
+                });
             }
             let pip_index_url = PIP_INDEX_URL
                 .read()
@@ -1189,7 +1265,7 @@ pub async fn handle_python_reqs(
 
             envs.push(("HOME", HOME_ENV.as_str()));
 
-            tracing::debug!("pip install command: {:?}", command_args);
+            tracing::debug!("uv pip install command: {:?}", command_args);
 
             #[cfg(unix)]
             {
@@ -1200,7 +1276,12 @@ pub async fn handle_python_reqs(
                     .envs(envs)
                     .args([
                         "-x",
-                        &format!("{}/pip-{}.lock", LOCK_CACHE_DIR, fssafe_req),
+                        &format!(
+                            "{}/{}-{}.lock",
+                            LOCK_CACHE_DIR,
+                            if no_uv_install { "pip" } else { "py311" },
+                            fssafe_req
+                        ),
                         "--command",
                         &command_args.join(" "),
                     ])
@@ -1211,16 +1292,20 @@ pub async fn handle_python_reqs(
 
             #[cfg(windows)]
             {
-                let mut pip_cmd = Command::new(PYTHON_PATH.as_str());
-                pip_cmd
-                    .env_clear()
+                let installer_path = if no_uv_install { command_args[0] } else { "uv" };
+                let mut cmd: Command = Command::new(&installer_path);
+                cmd.env_clear()
                     .envs(envs)
                     .envs(PROXY_ENVS.clone())
                     .env("SystemRoot", SYSTEM_ROOT.as_str())
+                    .env(
+                        "TMP",
+                        std::env::var("TMP").unwrap_or_else(|_| String::from("/tmp")),
+                    )
                     .args(&command_args[1..])
                     .stdout(Stdio::piped())
                     .stderr(Stdio::piped());
-                start_child_process(pip_cmd, PYTHON_PATH.as_str()).await?
+                start_child_process(cmd, installer_path).await?
             }
         };
 
@@ -1233,7 +1318,7 @@ pub async fn handle_python_reqs(
             false,
             worker_name,
             &w_id,
-            &format!("pip install {req}"),
+            &format!("uv pip install {req}"),
             None,
             false,
             occupancy_metrics,
@@ -1245,6 +1330,16 @@ pub async fn handle_python_reqs(
             "finished setting up python dependencies {}",
             job_id
         );
+        if child.is_err() {
+            
+            if let Err(e) = fs::remove_dir_all(&venv_p) {
+                tracing::warn!(
+                    workspace_id = %w_id,
+                    "failed to remove cache dir: {:?}",
+                    e
+                );
+            }
+        }
         child?;
 
         #[cfg(all(feature = "enterprise", feature = "parquet"))]
@@ -1253,7 +1348,7 @@ pub async fn handle_python_reqs(
                 tracing::warn!("S3 cache not available in the pro plan");
             } else {
                 let venv_p = venv_p.clone();
-                tokio::spawn(build_tar_and_push(os, venv_p));
+                tokio::spawn(build_tar_and_push(os, venv_p, no_uv_install));
             }
         }
         req_paths.push(venv_p);

backend/windmill-worker/src/result_processor.rs

@@ -295,6 +295,7 @@ pub async fn process_result(
                     message: format!("error during execution of the script:\n{}", err),
                     name: "ExecutionErr".to_string(),
                     step_id: job.flow_step_id.clone(),
+                    exit_code: None,
                 }),
             };
 
@@ -608,6 +609,8 @@ pub struct SerializedError {
     pub name: String,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub step_id: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub exit_code: Option<i32>,
 }
 pub fn extract_error_value(log_lines: &str, i: i32, step_id: Option<String>) -> Box<RawValue> {
     return to_raw_value(&SerializedError {
@@ -617,5 +620,6 @@ pub fn extract_error_value(log_lines: &str, i: i32, step_id: Option<String>) ->
         ),
         name: "ExecutionErr".to_string(),
         step_id,
+        exit_code: Some(i),
     });
 }

backend/windmill-worker/src/worker.rs

@@ -236,7 +236,14 @@ pub const TMP_LOGS_DIR: &str = concatcp!(TMP_DIR, "/logs");
 pub const ROOT_CACHE_NOMOUNT_DIR: &str = concatcp!(TMP_DIR, "/cache_nomount/");
 
 pub const LOCK_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "lock");
+// Used as fallback now
 pub const PIP_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "pip");
+
+// pub const PY310_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "python_310");
+pub const PY311_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "python_311");
+// pub const PY312_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "python_312");
+// pub const PY313_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "python_313");
+
 pub const UV_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "uv");
 pub const TAR_PIP_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "tar/pip");
 pub const DENO_CACHE_DIR: &str = concatcp!(ROOT_CACHE_DIR, "deno");
@@ -257,6 +264,7 @@ const NUM_SECS_PING: u64 = 5;
 const NUM_SECS_READINGS: u64 = 60;
 
 const INCLUDE_DEPS_PY_SH_CONTENT: &str = include_str!("../nsjail/download_deps.py.sh");
+const INCLUDE_DEPS_PY_SH_CONTENT_FALLBACK: &str = include_str!("../nsjail/download_deps.py.pip.sh");
 
 pub const DEFAULT_CLOUD_TIMEOUT: u64 = 900;
 pub const DEFAULT_SELFHOSTED_TIMEOUT: u64 = 604800; // 7 days
@@ -311,6 +319,7 @@ lazy_static::lazy_static! {
     .and_then(|x| x.parse::<bool>().ok())
     .unwrap_or(false);
 
+    // pub static ref DISABLE_NSJAIL: bool = false;
     pub static ref DISABLE_NSJAIL: bool = std::env::var("DISABLE_NSJAIL")
         .ok()
         .and_then(|x| x.parse::<bool>().ok())
@@ -740,6 +749,13 @@ pub async fn run_worker<R: rsmq_async::RsmqConnection + Send + Sync + Clone + 's
             "download_deps.py.sh",
             INCLUDE_DEPS_PY_SH_CONTENT,
         );
+
+        // TODO: Remove (Deprecated)
+        let _ = write_file(
+            &worker_dir,
+            "download_deps.py.pip.sh",
+            INCLUDE_DEPS_PY_SH_CONTENT_FALLBACK,
+        );
     }
 
     let mut last_ping = Instant::now() - Duration::from_secs(NUM_SECS_PING + 1);
@@ -1903,8 +1919,10 @@ async fn handle_queued_job<R: rsmq_async::RsmqConnection + Send + Sync + Clone>(
         }
     };
     if job.is_flow() {
+        let flow = job.parse_raw_flow();
         handle_flow(
             job,
+            flow,
             db,
             &client.get_authed().await,
             None,

backend/windmill-worker/src/worker_flow.rs

@@ -8,7 +8,6 @@
 
 use std::collections::hash_map::DefaultHasher;
 use std::collections::HashMap;
-use std::hash::Hash;
 use std::sync::atomic::{AtomicUsize, Ordering};
 use std::sync::Arc;
 use std::time::Duration;
@@ -49,7 +48,7 @@ use windmill_common::{
         Approval, BranchAllStatus, BranchChosen, FlowStatus, FlowStatusModule, RetryStatus,
         MAX_RETRY_ATTEMPTS, MAX_RETRY_INTERVAL,
     },
-    flows::{FlowModule, FlowModuleValue, FlowValue, InputTransform, Retry, Suspend},
+    flows::{has_failure_module, FlowModule, FlowModuleValue, FlowValue, InputTransform, Retry, Suspend},
 };
 use windmill_queue::schedule::get_schedule_opt;
 use windmill_queue::{
@@ -177,11 +176,6 @@ struct RecoveryObject {
     recover: Option<bool>,
 }
 
-#[derive(sqlx::FromRow, Deserialize)]
-pub struct RowFlowStatus {
-    pub flow_status: sqlx::types::Json<Box<serde_json::value::RawValue>>,
-    pub current_module: Option<sqlx::types::Json<Box<serde_json::value::RawValue>>>,
-}
 // #[instrument(level = "trace", skip_all)]
 pub async fn update_flow_status_after_job_completion_internal<
     R: rsmq_async::RsmqConnection + Send + Sync + Clone,
@@ -207,6 +201,7 @@ pub async fn update_flow_status_after_job_completion_internal<
     let (
         should_continue_flow,
         flow_job,
+        flow_value,
         stop_early,
         skip_if_stop_early,
         nresult,
@@ -215,35 +210,28 @@ pub async fn update_flow_status_after_job_completion_internal<
     ) = {
         // tracing::debug!("UPDATE FLOW STATUS: {flow:?} {success} {result:?} {w_id} {depth}");
 
-        let old_status_json = sqlx::query_as::<_, RowFlowStatus>(
-            "SELECT flow_status, raw_flow->'modules'->(flow_status->'step')::int as current_module FROM queue WHERE id = $1 AND workspace_id = $2",
+        let (old_status, current_module) = sqlx::query!(
+            "SELECT
+                flow_status AS \"flow_status!: Json<Box<RawValue>>\",
+                raw_flow->'modules'->(flow_status->'step')::int AS \"module: Json<Box<RawValue>>\"
+            FROM queue WHERE id = $1 AND workspace_id = $2 LIMIT 1",
+            flow, w_id
         )
-        .bind(flow)
-        .bind(w_id)
         .fetch_one(db)
         .await
-        .map_err(|e| {
-            Error::InternalErr(format!(
-                "fetching flow status {flow} while reporting {success} {result:?}: {e:#}"
-            ))
-        })?;
-
-        let old_status = serde_json::from_str::<FlowStatus>(old_status_json.flow_status.get())
-            .or_else(|e| {
-                Err(Error::InternalErr(format!(
-                    "requiring status to be parsable as FlowStatus: {e:?}"
-                )))
-            })?;
-
-        let current_module = if let Some(x) = old_status_json.current_module {
-            Some(serde_json::from_str::<FlowModule>(x.0.get()).or_else(|e| {
-                Err(Error::InternalErr(format!(
+        .map_err(|e| Error::InternalErr(
+            format!("fetching flow status {flow} while reporting {success} {result:?}: {e:#}")
+        ))
+        .and_then(|record| Ok((
+            serde_json::from_str::<FlowStatus>(record.flow_status.0.get()).map_err(|e| Error::InternalErr(
+                format!("requiring current module to be parsable as FlowStatus: {e:?}")
+            ))?,
+            record.module.map(|json| {
+                serde_json::from_str::<FlowModule>(json.0.get()).map_err(|e| Error::InternalErr(format!(
                     "requiring current module to be parsable as FlowModule: {e:?}"
                 )))
-            })?)
-        } else {
-            None
-        };
+            }).transpose()?,
+        )))?;
 
         let module_step = Step::from_i32_and_len(old_status.step, old_status.modules.len());
 
@@ -303,16 +291,15 @@ pub async fn update_flow_status_after_job_completion_internal<
 
             let is_flow = if let Some(step) = step {
                 sqlx::query_scalar!(
-                    "SELECT raw_flow->'modules'->($1)->'value'->>'type' = 'flow' FROM queue WHERE id = $2",
-                    step as i32,
-                    &flow
+                    "SELECT raw_flow->'modules'->($1)::text->'value'->>'type' = 'flow' FROM queue WHERE id = $2 LIMIT 1",
+                    step as i32, flow
                 )
-                    .fetch_one(db)
-                    .await
-                    .map_err(|e| {
-                        Error::InternalErr(format!("error during retrieval of step's type: {e:#}"))
-                    })?
-                    .unwrap_or(false)
+                .fetch_one(db)
+                .await
+                .map_err(|e| {
+                    Error::InternalErr(format!("error during retrieval of step's type: {e:#}"))
+                })?
+                .unwrap_or(false)
             } else {
                 false
             };
@@ -938,10 +925,7 @@ pub async fn update_flow_status_after_job_completion_internal<
             .unwrap_or_else(|| "none".to_string());
         tracing::info!(id = %flow_job.id, root_id = %job_root, "update flow status");
 
-        let module = get_module(&flow_job, &module_step);
-        // tracing::error!(
-        //     "UPDATE FLOW STATUS 3: {module:#?} {unrecoverable} {} {is_last_step} {success} {skip_error_handler} is_failure_step {is_failure_step}", flow_job.canceled
-        // );
+        let flow_value = flow_job.parse_raw_flow();
 
         let should_continue_flow = match success {
             _ if stop_early => false,
@@ -953,7 +937,14 @@ pub async fn update_flow_status_after_job_completion_internal<
             }
             false
                 if next_retry(
-                    &module.and_then(|m| m.retry.clone()).unwrap_or_default(),
+                    flow_value
+                        .as_ref()
+                        .and_then(|value| match module_step {
+                            Step::PreprocessorStep => value.preprocessor_module.as_ref().and_then(|m| m.retry.as_ref()),
+                            Step::Step(i) => value.modules.get(i).as_ref().and_then(|m| m.retry.as_ref()),
+                            Step::FailureStep => value.failure_module.as_ref().and_then(|m| m.retry.as_ref()),
+                        })
+                        .unwrap_or(&Retry::default()),
                     &old_status.retry,
                 )
                 .is_some() =>
@@ -963,7 +954,7 @@ pub async fn update_flow_status_after_job_completion_internal<
             false
                 if !is_failure_step
                     && !skip_error_handler
-                    && has_failure_module(flow, db).await? =>
+                    && has_failure_module(flow, db, false).await? =>
             {
                 true
             }
@@ -975,6 +966,7 @@ pub async fn update_flow_status_after_job_completion_internal<
         (
             should_continue_flow,
             flow_job,
+            flow_value,
             stop_early,
             skip_if_stop_early,
             nresult,
@@ -1042,13 +1034,11 @@ pub async fn update_flow_status_after_job_completion_internal<
                     let args_hash =
                         hash_args(db, client, w_id, job_id_for_status, &flow_job.args).await;
                     let flow_path = flow_job.script_path();
-                    let version_hash = if let Some(rc) = flow_job.raw_flow.as_ref() {
-                        use std::hash::Hasher;
-                        let mut s = DefaultHasher::new();
-                        serde_json::to_string(&rc.0)
-                            .unwrap_or_default()
-                            .hash(&mut s);
-                        format!("flow_{}", hex::encode(s.finish().to_be_bytes()))
+                    let version_hash = if let Some(sqlx::types::Json(s)) = flow_job.raw_flow.as_ref() {
+                        use std::hash::{Hash, Hasher};
+                        let mut h = DefaultHasher::new();
+                        s.get().hash(&mut h);
+                        format!("flow_{}", hex::encode(h.finish().to_be_bytes()))
                     } else {
                         "flow_unknown".to_string()
                     };
@@ -1107,6 +1097,7 @@ pub async fn update_flow_status_after_job_completion_internal<
         tracing::debug!(id = %flow_job.id,  "start handle flow");
         match handle_flow(
             flow_job.clone(),
+            flow_value,
             db,
             client,
             Some(nresult.clone()),
@@ -1239,19 +1230,6 @@ async fn retrieve_flow_jobs_results(
     Ok(to_raw_value(&results))
 }
 
-fn get_module(flow_job: &QueuedJob, module_step: &Step) -> Option<FlowModule> {
-    let raw_flow = flow_job.parse_raw_flow();
-    if let Some(raw_flow) = raw_flow {
-        match module_step {
-            Step::PreprocessorStep => raw_flow.preprocessor_module.map(|x| *x.clone()),
-            Step::Step(i) => raw_flow.modules.get(*i).map(|x| x.clone()),
-            Step::FailureStep => raw_flow.failure_module.map(|x| *x.clone()),
-        }
-    } else {
-        None
-    }
-}
-
 async fn compute_skip_branchall_failure<'c>(
     job: &Uuid,
     branch: usize,
@@ -1290,23 +1268,6 @@ async fn compute_skip_branchall_failure<'c>(
         }))
 }
 
-async fn has_failure_module<'c>(flow: Uuid, db: &DB) -> Result<bool, Error> {
-    sqlx::query_scalar::<_, Option<bool>>(
-        "SELECT raw_flow->'failure_module' != 'null'::jsonb
-        FROM queue
-        WHERE id = $1",
-    )
-    .bind(flow)
-    .fetch_one(db)
-    .await
-    .map_err(|e| {
-        Error::InternalErr(format!(
-            "error during retrieval of has_failure_module: {e:#}"
-        ))
-    })
-    .map(|v| v.unwrap_or(false))
-}
-
 // async fn retrieve_cleanup_module<'c>(flow_uuid: Uuid, db: &DB) -> Result<FlowCleanupModule, Error> {
 //     tracing::warn!("Retrieving cleanup module of flow {}", flow_uuid);
 //     let raw_value = sqlx::query_scalar!(
@@ -1523,6 +1484,7 @@ async fn transform_input(
 #[instrument(level = "trace", skip_all)]
 pub async fn handle_flow<R: rsmq_async::RsmqConnection + Send + Sync + Clone>(
     flow_job: Arc<QueuedJob>,
+    flow_value: Option<FlowValue>,
     db: &sqlx::Pool<sqlx::Postgres>,
     client: &AuthedClient,
     last_result: Option<Arc<Box<RawValue>>>,
@@ -1531,8 +1493,7 @@ pub async fn handle_flow<R: rsmq_async::RsmqConnection + Send + Sync + Clone>(
     rsmq: Option<R>,
     job_completed_tx: Sender<SendResult>,
 ) -> anyhow::Result<()> {
-    let flow = flow_job
-        .parse_raw_flow()
+    let flow = flow_value
         .with_context(|| "Unable to parse flow definition")?;
     let status = flow_job
         .parse_flow_status()

backend/windmill-worker/src/worker_lockfiles.rs

@@ -1312,6 +1312,8 @@ async fn python_dep(
             job_dir,
             worker_dir,
             occupancy_metrics,
+            false,
+            false,
         )
         .await;
 

benchmarks/benchmark_oneoff.ts

@@ -8,6 +8,7 @@ import { DenoLandProvider } from "https://deno.land/x/cliffy@v0.25.7/command/upg
 import { sleep } from "https://deno.land/x/sleep@v1.2.1/mod.ts";
 
 import * as windmill from "https://deno.land/x/windmill@v1.174.0/mod.ts";
+import * as api from "https://deno.land/x/windmill@v1.174.0/windmill-api/index.ts";
 
 import { VERSION, createBenchScript, getFlowPayload, login } from "./lib.ts";
 
@@ -172,8 +173,8 @@ export async function main({
       kind: "script",
       path: "f/benchmarks/" + kind,
     });
-  } else if (["2steps"].includes(kind)) {
-    nStepsFlow = 2;
+  } else if (["2steps", "bigscriptinflow"].includes(kind)) {
+    nStepsFlow = kind == "2steps" ? 2 : 1;
     const payload = getFlowPayload(kind);
     body = JSON.stringify({
       kind: "flow",
@@ -194,6 +195,15 @@ export async function main({
       kind: "script",
       path: kind.substr(7),
     });
+  } else if (kind == "bigrawscript") {
+    noVerify = true;
+    body = JSON.stringify({
+      kind: "rawscript",
+      rawscript: {
+        language: api.RawScript.language.BASH,
+        content: "# let's bloat that bash script, 3.. 2.. 1.. BOOM\n".repeat(25000) + "echo \"$WM_FLOW_JOB_ID\"\n",
+      },
+    });
   } else {
     throw new Error("Unknown script pattern " + kind);
   }
@@ -246,7 +256,7 @@ export async function main({
     } else {
       const elapsed = start ? Date.now() - start : 0;
       completedJobs = await getCompletedJobsCount();
-      if (kind === "2steps" || kind.startsWith("flow:")) {
+      if (nStepsFlow > 0) {
         completedJobs = Math.floor(completedJobs / (nStepsFlow + 1));
       }
       const avgThr = ((completedJobs / elapsed) * 1000).toFixed(2);

benchmarks/lib.ts

@@ -2,7 +2,7 @@ import { sleep } from "https://deno.land/x/sleep@v1.2.1/mod.ts";
 import * as windmill from "https://deno.land/x/windmill@v1.174.0/mod.ts";
 import * as api from "https://deno.land/x/windmill@v1.174.0/windmill-api/index.ts";
 
-export const VERSION = "v1.421.0";
+export const VERSION = "v1.424.0";
 
 export async function login(email: string, password: string): Promise<string> {
   return await windmill.UserService.login({
@@ -235,6 +235,23 @@ export const getFlowPayload = (flowPattern: string): api.FlowPreview => {
         ],
       },
     };
+  } else if (flowPattern == "bigscriptinflow") {
+    return {
+      path: "bigscriptinflow",
+      args: {},
+      value: {
+        modules: [
+          {
+            value: {
+              input_transforms: {},
+              language: api.RawScript.language.BASH,
+              type: "rawscript",
+              content: "# let's bloat that bash script, 3.. 2.. 1.. BOOM\n".repeat(25000) + "echo \"$WM_FLOW_JOB_ID\"\n",
+            },
+          }
+        ],
+      }
+    };
   } else {
     return {
       path: "2steps",

benchmarks/suite_config.json

@@ -30,5 +30,13 @@
   {
     "kind": "nativets",
     "jobs": 2000
+  },
+  {
+    "kind": "bigrawscript",
+    "jobs": 4000
+  },
+  {
+    "kind": "bigscriptinflow",
+    "jobs": 4000
   }
 ]

cli/main.ts

@@ -60,7 +60,7 @@ export {
 //   }
 // });
 
-export const VERSION = "1.421.0";
+export const VERSION = "1.424.0";
 
 const command = new Command()
   .name("wmill")

flake.nix

@@ -26,7 +26,10 @@
       ]);
     in {
       devShell = pkgs.mkShell {
-        buildInputs = buildInputs ++ (with pkgs; [ git go xcaddy sqlx-cli ]);
+        buildInputs = buildInputs ++ (with pkgs; [
+          git xcaddy sqlx-cli flock rust-analyzer
+          deno python3 python3Packages.pip go bun uv
+        ]);
         packages = [
           (pkgs.writeScriptBin "wm-caddy" ''
             cd ./frontend
@@ -56,6 +59,9 @@
             sqlx database create
             wm-migrate
           '')
+          (pkgs.writeScriptBin "wm-bench" ''
+            deno run -A benchmarks/main.ts -e admin@windmill.dev -p changeme $*
+          '')
           (pkgs.writeScriptBin "wm" ''
             cd ./frontend
             npm run dev $*
@@ -68,6 +74,12 @@
         DATABASE_URL = "postgres://postgres:changeme@127.0.0.1:5432/";
         REMOTE = "http://127.0.0.1:8000";
         REMOTE_LSP = "http://127.0.0.1:3001";
+        DENO_PATH = "${pkgs.deno}/bin/deno";
+        PYTHON_PATH = "${pkgs.python3}/bin/python3";
+        GO_PATH = "${pkgs.go}/bin/go";
+        BUN_PATH = "${pkgs.bun}/bin/bun";
+        UV_PATH = "${pkgs.uv}/bin/uv";
+        FLOCK_PATH = "${pkgs.flock}/bin/flock";
       };
       packages.default = self.packages.${system}.windmill;
       packages.windmill-client = pkgs.stdenv.mkDerivation {
@@ -82,7 +94,7 @@
           npm config set strict-ssl false
           cd frontend
           npm install --verbose
-          npm run generate-backend-client-mac
+          npm run ${if pkgs.stdenv.isDarwin then "generate-backend-client-mac" else "generate-backend-client"}
           NODE_OPTIONS="--max-old-space-size=8192" npm run build
         '';
 

frontend/package-lock.json

@@ -1,14 +1,15 @@
 {
 	"name": "windmill-components",
-	"version": "1.421.0",
+	"version": "1.424.0",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "windmill-components",
-			"version": "1.421.0",
+			"version": "1.424.0",
 			"license": "AGPL-3.0",
 			"dependencies": {
+				"@anthropic-ai/sdk": "^0.32.1",
 				"@aws-crypto/sha256-js": "^4.0.0",
 				"@codingame/monaco-vscode-configuration-service-override": "~8.0.2",
 				"@codingame/monaco-vscode-files-service-override": "~8.0.2",
@@ -21,6 +22,7 @@
 				"@codingame/monaco-vscode-standalone-typescript-language-features": "~8.0.2",
 				"@json2csv/plainjs": "^7.0.6",
 				"@leeoniya/ufuzzy": "^1.0.8",
+				"@mistralai/mistralai": "^1.3.0",
 				"@popperjs/core": "^2.11.6",
 				"@redocly/json-to-json-schema": "^0.0.1",
 				"@tanstack/svelte-table": "^8.9.9",
@@ -175,6 +177,30 @@
 				"node": ">=6.0.0"
 			}
 		},
+		"node_modules/@anthropic-ai/sdk": {
+			"version": "0.32.1",
+			"resolved": "https://registry.npmjs.org/@anthropic-ai/sdk/-/sdk-0.32.1.tgz",
+			"integrity": "sha512-U9JwTrDvdQ9iWuABVsMLj8nJVwAyQz6QXvgLsVhryhCEPkLsbcP/MXxm+jYcAwLoV8ESbaTTjnD4kuAFa+Hyjg==",
+			"license": "MIT",
+			"dependencies": {
+				"@types/node": "^18.11.18",
+				"@types/node-fetch": "^2.6.4",
+				"abort-controller": "^3.0.0",
+				"agentkeepalive": "^4.2.1",
+				"form-data-encoder": "1.7.2",
+				"formdata-node": "^4.3.2",
+				"node-fetch": "^2.6.7"
+			}
+		},
+		"node_modules/@anthropic-ai/sdk/node_modules/@types/node": {
+			"version": "18.19.64",
+			"resolved": "https://registry.npmjs.org/@types/node/-/node-18.19.64.tgz",
+			"integrity": "sha512-955mDqvO2vFf/oL7V3WiUtiz+BugyX8uVbaT2H8oj3+8dRyH2FLiNdowe7eNqRM7IOIZvzDH76EoAT+gwm6aIQ==",
+			"license": "MIT",
+			"dependencies": {
+				"undici-types": "~5.26.4"
+			}
+		},
 		"node_modules/@apidevtools/json-schema-ref-parser": {
 			"version": "11.6.1",
 			"resolved": "https://registry.npmjs.org/@apidevtools/json-schema-ref-parser/-/json-schema-ref-parser-11.6.1.tgz",
@@ -3659,6 +3685,14 @@
 			"resolved": "https://registry.npmjs.org/@mapbox/unitbezier/-/unitbezier-0.0.0.tgz",
 			"integrity": "sha512-HPnRdYO0WjFjRTSwO3frz1wKaU649OBFPX3Zo/2WZvuRi6zMiRGui8SnPQiQABgqCf8YikDe5t3HViTVw1WUzA=="
 		},
+		"node_modules/@mistralai/mistralai": {
+			"version": "1.3.0",
+			"resolved": "https://registry.npmjs.org/@mistralai/mistralai/-/mistralai-1.3.0.tgz",
+			"integrity": "sha512-G5DPCSC8sEhG3LUEZDYLD7qEZWDuZXgaX3IcoC3a/ydm9jFuh2pRZtknsgMx2sU8d7kxRuxblY3fPH5C38wnhQ==",
+			"peerDependencies": {
+				"zod": ">= 3"
+			}
+		},
 		"node_modules/@nodelib/fs.scandir": {
 			"version": "2.1.5",
 			"resolved": "https://registry.npmjs.org/@nodelib/fs.scandir/-/fs.scandir-2.1.5.tgz",
@@ -13738,6 +13772,16 @@
 			"integrity": "sha512-A7AMeJfGefk317I/3tBoUYRcDcNavKEkpiPN/nQsBz/viI2GvT7BtrqdPD6rGqBFN8Ax7v4obf+Cl32JF9DDVw==",
 			"dev": true
 		},
+		"node_modules/zod": {
+			"version": "3.23.8",
+			"resolved": "https://registry.npmjs.org/zod/-/zod-3.23.8.tgz",
+			"integrity": "sha512-XBx9AXhXktjUqnepgTiE5flcKIYWi/rme0Eaj+5Y0lftuGBq+jyRu/md4WnuxqgP1ubdpNCsYEYPxrzVHD8d6g==",
+			"license": "MIT",
+			"peer": true,
+			"funding": {
+				"url": "https://github.com/sponsors/colinhacks"
+			}
+		},
 		"node_modules/zstddec": {
 			"version": "0.1.0",
 			"resolved": "https://registry.npmjs.org/zstddec/-/zstddec-0.1.0.tgz",