* feat: ask before minting the frontend sdk token in the raw app editor
* fix: record editor sdk consent on continue so a late mint cannot override a decline
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: only the newest editor sdk mint may start the preview
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat: pause a worker group without stopping its workers
* chore: bump ee ref
* fix: keep dedicated workers running across a pause toggle
* fix: edit a copy of the worker group config in the drawer
* fix: stable priority tag order, keep pause-only configs in the drawer
* fix: keep inherited worker tags when saving from the drawer
* fix: keep an explicit empty tag list when leaving dedicated mode
* fix: keep inherited tags when creating a config from the drawer
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore: update ee-repo-ref to 6f23819e947b18878000e580ddefccfd01a73166
This commit updates the EE repository reference after PR #853 was merged in windmill-ee-private.
Previous ee-repo-ref: 6cf802bdbe17b4a66757016515044b8a9ef9ef19
New ee-repo-ref: 6f23819e947b18878000e580ddefccfd01a73166
Automated by sync-ee-ref workflow.
* feat: confirm before applying a worker config change that restarts workers
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore: pin the ee commit that compiles against main
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore: retrigger ci after the github outage dropped the push event
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com>
* fix: type snowflake columns in database manager insert and alter forms
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: scope insert-form type lowercasing to snowflake
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: send snowflake insert-form dates as yyyy-MM-dd
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore: update ee-repo-ref to acb254ac509aff244849c09e50a8d39aae185138
This commit updates the EE repository reference after PR #852 was merged in windmill-ee-private.
Previous ee-repo-ref: 9cc330f1a8e80bc075a6ad9839966d3b344f8ddd
New ee-repo-ref: acb254ac509aff244849c09e50a8d39aae185138
Automated by sync-ee-ref workflow.
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com>
Co-authored-by: Ruben Fiszel <ruben@windmill.dev>
* feat: run AI decision steps on OpenAI's Decisions API
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: fail refused openai decisions and keep structured criteria
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: send only user message lists as openai decision messages
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: omit descriptions the openai decision step does not have
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: pass only exact openai user messages through as decision input
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: check the payload of openai decision message parts
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: default new raw apps to react 19 and ask for the data setup
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: say what app new --schema writes and soften the already-set-up wording
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: resolve bare init_app table names in the app schema, fix --schema help
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: make every trigger and schedule editor control respect read-only mode
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: grey out read-only option buttons in gcp, mqtt and sqs editors
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix: keep chat tool cards collapsed the way the user left them
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: open a tool card awaiting confirmation even after the user collapsed it
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: set up a new app before starting with AI and follow its deploy
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: set up a new app in one modal over its ai session
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: show schema creation progress and put the new app modal over the page
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: say when no worker can create a new app's schema instead of waiting
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: drop the summary from the new app modal
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: list existing tables only under an existing schema, explore from the schema row
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: start a new app from two cards, with ai recommended
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: isolate the raw-app editor preview when the app opts into sandboxing
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: gate job reads from a sandboxed editor preview like the deployed app
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: send a plain user in a sandboxed raw app's startup message
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: pin ui_builder to d3fdb70
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat: choose build with AI or code when creating a raw app
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: carry the picked data table into a new-app session and seed only on 404
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: wait for data tables before build with AI and stop re-registering on edits
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: create apps without an unusable data table and name session-seeded apps
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: rename new-app choices and hint where to switch to AI later
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: open a run as a pushed page on narrow runs page
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: keep modifier-click multi-select and arrow keys on the narrow runs list
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: leave the narrow re-run page once the re-run is sent, and clear the opened run on reset
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: close the re-run options once a table re-run completes
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: serve the raw-app preview shell sandboxed, load a blob copy in the editor
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: give the preview shell the deployed app wrapper's sandbox flags
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix: drop the add a prompt for AI button from the run page
* fix: open the run page in the session preview from a plain chat button
* fix: rename additional prompt for AI to AI prompt
* fix: put the run page chat button beside advanced when no AI prompt
* fix: search slow runs filters by time windows instead of timing out
* fix: cancel a pending runs page when its listing is reset
* fix: keep runs scan windows disjoint at microsecond precision
* fix: race the single runs request against the window scan
* fix: list the cursor window when a runs scan starts at the oldest job
* fix: show the runs history scan as an info alert
* fix: show runs scan progress as a bar and keep the session button on the searchbar row
* fix: stop a runs scan whose windows are no faster when narrower
* fix: right-align the runs searchbar row when the header wraps
* fix: keep the runs scan stop button inside a narrow list pane
A team plan canceled mid-month now keeps premium until the 1st, capped at the
executions its final invoice paid for. That reuses the past-due cap, so the job
refusal and the UI toast said "your last invoice is unpaid". Word both for a
canceled plan, and show it on the plan settings page.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(ai-chat): show per-file diffs for raw app edits
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: keep reference runnable writes on the generic tool card
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: extend diff row highlights across horizontal overflow
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: collapse queued and running chat tool calls into a waiting row
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: correct the waiting-row comment in groupToolRuns
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: fold queued calls only behind a call in flight, after edit groups
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: stop a waiting row before queued calls that form an edit or explore group
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* refactor: use the session tab strip for workflow sdk module tabs
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: keep module deletion on the explicit button and tab order view-only
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* refactor: use the design-system button for module deletion
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: let modified keys through file tabs and keep keyboard rename off the tab
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: route modified enter past the dnd zone so page shortcuts fire from draggable tabs
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: forward modified enter from dnd item wrappers too
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: evaluate field conditions and step test inputs in QuickJS
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: answer single-quoted field conditions without loading the engine
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: bound expression recursion and ignore leading blank space
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat: card layout and load-more for the audit logs page
* fix: stream audit log load-more with stop control and show inline filters from fhd
* fix: fit audit log filters to measured header width and hold timeline during load-more
* fix: keep audit log load-more available when a trimmed short batch has rows left
* fix: stream the ai chat compaction request instead of waiting on it
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: reject a compaction stream stopped midway instead of keeping partial text
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: reject a compaction stream that ends without a finish reason
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test: cover a completed responses stream in the streamed completion helper
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat: redesign the service logs page
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: drop a stale jump after a cached host switch and remove unused max_lines
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: keep focus on url sync and expose host selection and new logs to assistive tech
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: reset the tail on dropdown host picks, resume refresh after search, key log cache by host
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: clear the pending jump when a host switch needs no fetch
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: show upcoming events when hovering a schedule's cron expression
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: read cron version from the draft for draft-only schedule previews
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: replace the AI input filling toggle with an additional prompt for AI
* feat: pass the additional prompt for AI to MCP clients and shared AI guidance
* feat: shrink the run page AI card to a button and a collapsed prompt
* fix: offer writers a way to add a prompt for AI from the run page
* fix: let legacy draft-only items be discarded from the home page
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: describe the legacy draft move guard as it now is
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: add an instance setting routing all dependency jobs to one tag
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: keep bunnative locks on bun and explain the default dependency routing
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* feat: add test key button to AI resource drawers
* fix: scope-check inline AI resource values and keep test model editable
* fix: keep test model editable for unsaved resources, own-key provider check
* fix: restore the save to workspace button on inline flow steps
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: make the inline step save to workspace button icon only
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: offer user scopes in the slack scope editor and drop the generated description on type change
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat: offer slack write scopes as bot and user scope options
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(sessions): show an item's deployed page in the preview panel
The preview panel showed an item's editor and nothing else. It now shows the
deployed page too, as a tab of its own.
**Two sides, two tabs.** The editor and the deployed page are different things
to look at, so `isItemTabFor` keys on the side rather than the item: asking for
a side already open focuses it, asking for the other opens it alongside. Two
editors for one item still cannot coexist — they would race its single
(kind, path) cell — but an editor beside a viewer is safe, since the viewer
reads the deployed version over the API and holds no cell. The tab breadcrumb
keeps re-pointing in place; `Exit & see details` and the detail page's `Edit`
open the other side instead of consuming the one you are on. A tab's label
carries `(edit)` when it is the editor side.
**The detail pages moved out of their routes.** `/scripts/get` and `/flows/get`
are now thin wrappers over `ScriptDetail` and `FlowDetail`, which the panel
renders too. Everything they reach — drawers, triggers, saved inputs — is scoped
to the viewer's workspace through `setOperatingWorkspace`, and every navigation
they attempt is caught by `interceptNav` and turned into a move inside the
panel, so nothing takes the browser out of the session. A plain click is
intercepted; ⌘-click and middle-click still open a real tab, which is why the
pages keep their `href`s.
**The picker opens what exists.** A row opens the deployed page, or the editor
when nothing is deployed there, and carries a Draft / Draft only badge in the
review dock's words. `WorkspaceItem` gained `draftOnly` and `hasDraft` from the
listers, which already returned both; deploys now invalidate the picker cache
through `itemDeployed`, so a just-deployed item stops reading as a draft. Rows
also carry a hover Edit action, reachable from the keyboard with the pick
modifier.
`open_preview` gained a `mode`, narrowed out of the advertised schema for a
session that cannot write drafts and re-checked per path in the handler, where
a refusal reports "couldn't check" apart from "denied". The mode is resolved
before those checks, so a call that omits it is gated as the editor it will
become.
Alongside, the workspace a detail page acts on is now the one it is showing
rather than the one the browser is navigated to: `MoveDrawer`,
`toggleWorkspaceErrorHandler` and the pages' own permission gates read the
operating workspace and its acting user, and `RunForm` mints a password
argument's ephemeral variable there too — in a fork session all of these
previously answered for the parent. `InWorkspaceAppViewer` hands the app's
`ctx` user down as a prop instead of writing the global `userStore`, which from
a session tab was re-pointing every permission check on the page.
Fixes found on the way: `DetailPageLayout` claimed `h-screen` inside a panel
that is not the viewport; Edit on a historical script version dropped the
version from the intercepted click; the Run button stayed spinning after a run
that left the page mounted; and the window-level run shortcut fired from a
collapsed panel and from keys another handler had already claimed.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* fix(sessions): historical edits, stale not-found, shared edit rights, tab labels
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): keep deployed-page links and workspace reads in the session
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): scope detail hrefs to the session, re-point the viewer's own tab
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): keep a previewed raw app's route out of the session URL
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): resolve edit-in-fork against the session workspace
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): read advanced run tags from the session workspace
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): fork from the session workspace, star only navigation items
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): mount only the side a preview tab shows
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): address CI review round 1 findings
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): route links in a deployed page's drawers through the panel
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): tell the chat which deployed page the panel shows
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): only a 404 reads as an undeployed raw app
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): scope the unparseable-JSON run gate to the form
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): a flow deleted from the panel stays in its tab
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): the picker's edit shortcut works on the current row
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): load-error toasts say what failed
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(sessions): keep an oversized invalid JSON editor in its form's run gate
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* revert: keep main's draft_only description in openapi.yaml
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
* feat: let slack connects issue a user token via user_scope
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: keep trailing newline in ee-repo-ref
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: bump ee-repo-ref
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: name the provider in slack token descriptions and refresh them on reconnect
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: drop the generated slack description when connecting another provider
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: drop the slack scope pin migration, slack ignores scope on refresh
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix: clear the generated slack description on client-credentials connects too
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* chore: update ee-repo-ref to ff4d04f17721d84fb9a1f655af50ff7e4e7dcbc7
This commit updates the EE repository reference after PR #837 was merged in windmill-ee-private.
Previous ee-repo-ref: 83298dcf23574d5ed05e6c767bf1d9b067ab7a95
New ee-repo-ref: ff4d04f17721d84fb9a1f655af50ff7e4e7dcbc7
Automated by sync-ee-ref workflow.
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Ruben Fiszel <ruben@windmill.dev>
Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com>
* feat: show human labels for oauth scope options in the connect dialog
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat: always show scope checkboxes, label gdocs and gchat scopes
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat: offer default scopes as checkboxes for every oauth provider
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix: block runs while a JSON input does not parse
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: clear json editor error on unmount and flush editors before run
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: restore validity when a nullable arg input is cleared
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: let field editors parse before the run check and reset the message on view switch
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix: clear the run refusal message when form validity changes
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* oauth: add gdocs and gchat providers, gchat icon
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* oauth: Google sign-in button for every accounts.google.com provider, least-privilege gchat default
- AppConnectInner: derive isGoogleSignin from the registry auth_url instead of a
hardcoded list, so gdocs/gchat/gforms/gcloud/gworkspace get Google's button.
- gchat default scopes: chat.messages is a restricted scope; default to
chat.spaces.readonly + chat.messages.create (both sensitive). chat.messages /
chat.messages.readonly stay selectable.
- BRAND_COLORS.md: GchatIcon row in sort order.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat: make hub integrations usable as examples in global chat
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* fix: keep web search guidance in sync with provider capability
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* fix: resync web search guidance before the request is built
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* fix: carry loop web search availability into every prompt rebuild
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* fix: drop web search guidance on the completions api fallback
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* docs: correct onBeforeIteration contract for the fallback re-entry
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* feat: give the ai chat hub script descriptions and integration metadata
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: treat authored hub metadata as evidence the scripts were curated
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: treat hub curated as three-state and speak only for a stated true
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* test: let the benchmark hub serve integration metadata
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* test: benchmark the hub tool against a real integration's scripts and metadata
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* test: cover two more real integrations where the scripts already answer
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* test: report metadata_source from whether the fixture has authored meta
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: parse the hub resource type schema instead of relaying it as a string
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: match integration suggestions on slug words instead of substrings
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: search the integration a query names outright instead of ranking past it
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: reject hub integration slugs that would re-target the proxied request
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: add a mentioned integration's hits instead of filtering the search to it
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: keep named-integration hits under the content cap and unmangle fixture placeholders
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: cut the middle of a capped hub result instead of repeating its tail
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: cap hub results by keeping the best of the ranked and named hits
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: keep a named integration that ranking placed below the content cap
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: let plan mode use the hub integration lookup
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* feat: mark which hub integrations carry authored provider knowledge
The hub flags the integrations whose document holds hand-written provider
knowledge, so a caller can tell before spending a call on the metadata
endpoint: 18 of ~216 qualify, and for the rest the endpoint returns what
was inferred from the same scripts a search already hands back.
Carry the flag onto search results, where the model first meets a slug, so
get_hub_integration is aimed at the few instead of guessed at. A hub that
predates the flag omits it and nothing is marked.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: stop the documented mark reading as a reason not to call
The flag says the hub additionally holds provider knowledge checked against
the live API. It is not a signal to skip the lookup elsewhere: that call
still returns the resource type and the usage-ranked examples, neither of
which a search result carries, and neither guessable.
The prompt said to read a script instead when the mark is absent, which
traded those for a guess on the ~198 undocumented integrations.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: keep the benchmark's documented flag true to what it serves
The eval hub reported Baremetrics and Holded as undocumented while its
metadata endpoint handed back their authored notes, so a case could teach
the model the flag means nothing. Derive it from both fixture sources, and
pin the agreement.
Also trims the documentedIntegrations comment to the four lines AGENTS.md
allows, keeping the case-folding constraint.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: keep the benchmark hub mostly undocumented, like the real one
Adding holded and baremetrics to the documented set resolved the fixture's
contradiction the wrong way: it left five of six integrations marked, against
the live hub's 18 of ~216, and claimed authored notes for two the hub reports
as having none. Drop the notes instead. Their auth and endpoints are in their
shipped scripts, which is what the cases that use them are about.
Also drops the last two places still describing the flag as a reason to spend
or skip the metadata call.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* refactor: let the model name the integration instead of guessing it from the query
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* refactor: word web search guidance conditionally instead of tracking provider capability
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: open the hub metadata route to job tokens and share one integrations fetch
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: record a copied hub script's source where write_script keeps it
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* test: pin that a query narrows to the integration it was given
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: match integration suggestions on the name the hub curates
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix: name an integration the hub leaves unnamed from the local word table
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
GuilhemandClaude Opus 5.5
AlexRV12andClaude Opus 5.5
hugocasaandClaude Opus 5.5