mirror of
https://github.com/windmill-labs/windmill.git
synced 2026-08-26 00:01:37 +00:00
a015ae00a5de8a8efc6d689b153e019d6bdbe3c1
14517 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 hugocasa
|
a015ae00a5 |
fix: let plan mode use the hub integration lookup
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
7667de3b5f |
Merge remote-tracking branch 'origin/main' into hub-integrations-prominence
# Conflicts: # ai_evals/adapters/frontend/core/global/globalEvalRunner.ts # ai_evals/adapters/frontend/mockBackend.ts # ai_evals/cases/global.yaml # frontend/src/lib/components/copilot/chat/AIChatManager.svelte.ts # frontend/src/lib/components/copilot/chat/global/core.ts |
||
|
hugocasa
|
532daa82d6 |
fix: keep a named integration that ranking placed below the content cap
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
66f204c7b8 |
fix: cap hub results by keeping the best of the ranked and named hits
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
03d06db7a2 |
fix: cut the middle of a capped hub result instead of repeating its tail
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
aa51731ad1 |
fix: keep named-integration hits under the content cap and unmangle fixture placeholders
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
4ed27eb62b |
fix: add a mentioned integration's hits instead of filtering the search to it
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
7983bdf5a7 |
fix: reject hub integration slugs that would re-target the proxied request
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
bc0bef14c2 |
fix: search the integration a query names outright instead of ranking past it
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
 Ruben Fiszel
|
0f3d884c6f |
chore(main): release 1.796.0 (#10810)
* chore(main): release 1.796.0 * Apply automatic changes --------- Co-authored-by: rubenfiszel <275584+rubenfiszel@users.noreply.github.com>v1.796.0 |
||
 Diego Imbert
|
3b2a6d7604 |
feat(datatables): add a down migration from the migration viewer (#10812)
* feat(datatables): add a down migration from the migration viewer Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ha889ogajX9jbqmwTaF8kD * fix(datatables): refuse an empty down migration and keep the saved one visible Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ha889ogajX9jbqmwTaF8kD * refactor(datatables): use unifiedSize on the new buttons and fix the lock comment Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ha889ogajX9jbqmwTaF8kD * fix(datatables): make the add-down exemption atomic against concurrent additions Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ha889ogajX9jbqmwTaF8kD * fix(datatables): re-test the whole observed row when an upsert skips the lock Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ha889ogajX9jbqmwTaF8kD * fix(datatables): re-test the observed row even when none was read, and sync the down draft on the leading change Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ha889ogajX9jbqmwTaF8kD --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
2906504125 |
feat: add instance setting to mute zombie job restart alerts (#10813)
* feat: add instance setting to opt out of zombie job restart alerts Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: preserve explicit false for default-on boolean instance settings Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * refactor: invert zombie restart alert setting to a mute flag Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
 AlexRV12
|
93081e255f |
fix(frontend): mint string password secrets in the operating workspace (#10815)
* fix(frontend): mint password secrets in the operating workspace A `password: true` string argument is rendered by PasswordArgInput, which mints an ephemeral secret variable on the first keystroke and rebinds the argument to `$var:<path>`. It minted into `$workspaceStore` — the globally active navigation workspace. Session editors operate on a different, possibly forked workspace without switching `$workspaceStore`, and thread that operating workspace explicitly as a `workspace` prop. When the two diverged the secret landed where the user was merely looking while the job ran elsewhere, and the backend failed with `Variable not found`. Add the `workspace` prop to PasswordArgInput and thread it through every hop between a form mount and the minting field, plus the entry points that supply it. Track `mintedIn` so updates target where the variable actually lives, and re-mint when the operating workspace moves after a path already exists. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(frontend): keep a password field consistent with its argument A parent can replace the whole args object without remounting this field — previewing a saved input, say — leaving `path` and `password` describing a secret the argument no longer points at. Minting from them then copies the old plaintext over the replacement, and the replacement is lost. State that rule once as `argReplaced` and gate every mint on it. The replacement can also land while the create is in flight, so the bound value is captured before the request and re-checked after it resolves; the variable that mint produced was never referenced, so it is deleted outright. A mint that ends without binding re-seeds `password` from what the argument now holds, so the field stops displaying a secret that will not be submitted and a later workspace move cannot re-mint the stale plaintext. `updateValue` returns early before anything is minted, since its 404 retry would otherwise bind over a replacement it cannot see. A failed initial mint now raises a toast rather than passing silently, which also removes the component's last unhandled rejection. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * test(frontend): guard workspace forwarding to PasswordArgInput Every hop between the form a caller mounts and the PasswordArgInput that mints the secret must forward `workspace`, and so must the entry points that supply it. A hop that drops the prop falls back to the navigation workspace while the top-level case keeps passing, and no typechecker catches it because every hop declares `workspace?: string | undefined`. The forwarded expression is checked rather than the prop's presence, so `workspace={$workspaceStore}` and `workspace={undefined}` fail. Two ways the scan could stop guarding without failing are asserted too: an unterminated mount raises instead of swallowing the rest of the file, and the number of mounts parsed must equal the number of tag occurrences, so a mount written inline rather than at the start of a line fails loudly. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(frontend): list and create variables in the operating workspace * fix(frontend): surface and bound a failed recovery mint * test(frontend): end a mount at the first line closing it --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Diego Imbert
|
25a3e6ea7a |
fix(ai-chat): keep the composer usable while a question is pending (#10816)
* fix(ai-chat): keep the composer usable while a question is pending Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01D2hyAfRT2aFF7uswsdTodL * fix(ai-chat): keep a typed answer when the question's resolver is gone Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01D2hyAfRT2aFF7uswsdTodL * fix(ai-chat): only advertise the answer affordance on a live question Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01D2hyAfRT2aFF7uswsdTodL * style: trim the pending-question rationale comments to the 4-line cap Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01D2hyAfRT2aFF7uswsdTodL --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
 Guilhem
|
29c311ab31 |
fix: qualify foreign key targets in generated datatable migrations (#10821)
* fix: qualify foreign key targets in generated datatable migrations The schema API reports a foreign key's target as a bare table name when it lives in the same schema as the table declaring it. Emitted verbatim that becomes `REFERENCES tickets (id)`, which Postgres resolves against search_path — and a migration's own schema is never on it, so applying it fails with `relation "tickets" does not exist` and the whole transaction rolls back. Nothing is created; the project imports with no tables. qualifyFkTarget resolves the target the way the FK closure does: the declaring table's schema first, then any schema holding that table. The REFERENCES clause is now quoted per part, so a qualified target survives identifiers that need quoting; the constraint name is still built from the unquoted value, so the pg_constraint guard still matches what it creates. `quoteTarget` is opt-in, so alterTable.ts — the only other caller of renderForeignKey — is byte-identical. Reproduced and verified against a real data table: hub.windmill.dev's published helpdesk migration fails as above, and the same SQL with the target qualified creates both tables and the constraint. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Xg8vXUuHCH3aRkf91sfxjx * fix: resolve bare foreign key targets in the declaring schema Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Wq66AbLqo4c5ukWJSc4x4t --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
541b6c8496 |
fix: keep ai chat messages when leaving the page mid-generation (#10809)
* fix: persist ai chat turns mid-generation so leaving the page keeps them * fix: stop chat checkpoints once the turn commits, keep streamed text visible * fix: checkpoint streamed answers as they grow and keep half-run tool batches * fix: checkpoint text as received so a backgrounded tab keeps capturing * fix: keep buffered tool screenshots in mid-batch chat checkpoints * fix: decide committed-text at the flush site, condense checkpoint comments * fix: checkpoint only live streamed text, never text the parser owns * fix: don't swap the chat transcript out from under a running turn * fix: close the pre-loading window in the conversation-switch guard |
||
|
Ruben Fiszel
|
8dbd12ecc1 |
fix: patch sqlx so a cancelled BEGIN cannot poison a pooled connection (#10823)
* fix: patch sqlx so a cancelled BEGIN cannot poison a pooled connection Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Qzqmh52NU8fB9RBQNNkJGt * test: drop the migration run and fixed sleep from the sqlx patch guard Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Qzqmh52NU8fB9RBQNNkJGt * test: ignore the sqlx patch guard by default and point at it from where sqlx is changed Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Qzqmh52NU8fB9RBQNNkJGt --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
3d47bf8caa |
fix: match integration suggestions on slug words instead of substrings
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
9c557859c5 |
feat: AI agent evals: datasets, scored runs and comparison (#10633)
* feat: eval datasets and standalone runs for reusable AI agents Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: agent eval drawer with case editor, runs and capture entry points Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * docs: document AI agent eval datasets and standalone runs Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: say how many eval cases the list is not showing Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address review findings on eval datasets - keep an edited case's conversation and tool inputs: serde(flatten) silently drops Box<RawValue> fields, so the update payload is spelled out - remount the case editor per case so one case's turns cannot leak into another - require jobs:read / flow_conversations:read on the capture endpoints, which UserDB does not gate by token scope - take the dataset lock in create and update so a delete cannot be undone by a concurrent metadata write, and delete cases before metadata - load more cases beyond the first page, and stop capping the agent picker - record that the version stamp is taken at enqueue, not at resolution Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address round-2 review findings on eval datasets - block operators from dataset and case writes - pass the editor's operating workspace through the drawer and the capture request, instead of assuming the navigation workspace - discard superseded case-list responses so switching datasets cannot land the previous dataset's cases - reject a dataset without a case_id (or vice versa) rather than running an inline case under a dangling association - run unsaved edits inline instead of silently running the stored case - surface the API error body on a failed run - fetch dataset metadata concurrently when listing - $bindable() without a default on the optional open prop - correct the permission and enqueue-time-version wording in the docs Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: run an untouched saved case by reference again The editor writes back keys the stored case omits, so comparing the raw objects reported every unedited case as edited: the run went inline and lost the dataset/case stamp its history depends on. Compare a normalized form, and pin it with a test. Also scope the history query to the drawer's workspace and drop superseded responses. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: show a dataset's cases as a table, and fix round-4 review findings The case list showed one case at a time with no overview. It is now a table with the case, where it was captured from, and its last run — the last-run column is a single jobs query on the path stamp rather than a request per row. Review fixes in the same file: - keep the edit baseline on the selected case rather than looking it up in the loaded page, so a case beyond page 1 is not treated as unedited and run stale - release the loading state when a superseded case load returns early - reload every loaded page after a write instead of collapsing to page 1 - last remaining 'resolved to' wording in the version tooltip Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: run a dataset as an experiment, with scorers as runnables An experiment runs every case of a dataset against one subject and records the exact case set it executed, so a result set stays reproducible while the dataset keeps changing. Each case runs as its own small flow — the agent, then a step per scorer — so a case keeps the run stamp, history query and trajectory view a single run already has, and scorers need no orchestration of their own. Results are read back per step by node id rather than by walking a nested loop's status. A scorer is any runnable taking (input, output, expected): a script, a flow, or a reusable agent used as a judge. A judge is prompted with the case and the answer as one JSON message; a script or flow receives them as named arguments. Scores accept a bare number, a boolean or {score}. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: results table for an experiment, with scorer columns One row per case: status, the agent's answer, and a column per scorer, with the mean per scorer above the table and a link into each case's run for its trajectory. Averages skip cases a scorer produced no number for — counting a missing score as zero would read as a regression. The drawer's left pane becomes Cases / Results, and Results carries the scorer picker and Run dataset. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: compare an experiment against a baseline Per-scorer deltas on each row and on the mean, and a filter down to the rows that regressed. Rows join by case id, so a case added after the baseline ran has no delta instead of counting as a change. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address round-5 review findings on experiments - match scorers by label when diffing two experiments; joining by array position subtracted one scorer from another whenever the scorer sets differed - report a row's status from the case job, not the agent step, so a case whose scorer failed no longer reads as a success - delete a dataset's experiments with it: they hold copies of its cases, and a recreated dataset of the same path would have exposed them - select the experiment that Run dataset just started instead of leaving the table on the previous one - expected is scored now, so stop describing it as having no consumer Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address round-6 review findings on experiments - hold the dataset lock across an experiment launch, so a delete landing between reading the cases and writing the experiment cannot recreate the deleted dataset's inputs - match scorers between experiments on kind and path, not on label: labels default to a path's last segment, so f/a/quality and f/b/quality compared against each other - average mean deltas over the cases both runs scored; comparing each run's own average reported a regression from a case the baseline never ran, with no regressed row to point at - openapi: the row status is the job's, which is also canceled/skipped; runEval takes scorers; the update-case body no longer advertises source, which the handler deliberately ignores - record why the experiment prefix cannot reach a sibling dataset Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address round-7 review findings on experiments - release the dataset lock for the push loop and retake it for the write, re-checking the dataset still exists: holding it across the whole launch made every capture and case edit on that dataset 409 until the last job queued - assemble experiment results with bounded concurrency; a 100-case, 3-scorer experiment was 400 sequential lookups, each itself several queries - clear the baseline when it becomes the selected experiment, which was comparing a run against itself and reporting zero deltas - take the header mean over the same cases as its delta while comparing, so the two numbers beside each other describe the same set - a canceled or skipped case is no longer the same grey dot as a running one Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address round-8 review findings on experiments - verify the dataset's identity, not just its existence, before recording an experiment: the path can be deleted and recreated during the push loop, and the experiment holds copies of the old dataset's cases - give the recording lock a longer budget than a case edit, since its jobs are already queued and giving up strands them, and say so when it fails - keep score lookups sequential within a case: nesting two bounded streams multiplied into 32 in-flight queries against a 50-connection pool - clear a baseline that no longer belongs to the loaded experiments, so switching datasets does not leave comparison mode on with nothing to compare - keep a scorer's own mean when the baseline never ran it, instead of blanking a column full of numbers - EvalCaseDraft.expected no longer claims nothing scores it Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: do not trust an experiment's job ids, and require write to record one Experiment objects live in workspace object storage, which a script can write directly, and results are read on the unrestricted pool — so a forged experiment naming another flow job returned output the jobs API would have refused. Only jobs this server stamped with that experiment's id are read now. Also from round 9: - recording an experiment requires write on the dataset, not read: it persists into the dataset's namespace and its shared list - clear the results table when the selection changes and surface a failed load, instead of labelling the previous experiment's numbers as the new one's - a storage fault is no longer reported as a deleted dataset - the lock-timeout message at the recording site no longer says to retry, which would run the whole dataset again on top of the jobs already queued - ExperimentRow.status documents canceled and skipped Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: bind the experiment trust check to the requested dataset The previous check matched jobs on the experiment id alone, which the stored object supplies — so copying another dataset's experiment JSON under a readable key carried its jobs' output along with it. A job is now only read if it was stamped for this experiment *and* for the dataset the caller's read access was checked against, and an experiment that names a different dataset is not served from this key at all. Also from round 10: - add the .sqlx entry for that query; without it every SQLX_OFFLINE build failed - serve results over GET: as POST the route-scope middleware classified a read as ai_evals:write, locking read-only tokens out of their own results - clear the selected and baseline experiments synchronously when the dataset changes, so the previous dataset's id is not requested under the new one Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address round-11 review findings on experiments and scorers - give scorers the whole case input, not just the message: an answer that came from attachments or a replayed conversation could not be judged on it - accept a judge's boolean and structured {score} answers, including stringified ones, and pin every documented scorer shape with a test - record an experiment for the cases that did launch when a later push fails, instead of leaving those jobs running with nothing to attribute them to - do not capture a preview parent's synthetic runnable_path as a host flow; the saved case could not be rerun - clear the case table before loading a dataset and surface a failed load, so a failure cannot leave the previous dataset's cases under the new name - keep the results table through a refresh of the same experiment - exclude flow-step jobs from the per-case last-run lookup - drop case sets from the experiment list, which is only used to pick a run - report a database failure at the recording lock as itself, not as contention Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address round-12 review findings on capture and run history - load flow_node.flow for flownode parents: an agent inside a deployed branch or loop captured without its agent, host flow or tool bindings - decide host_flow_path by whether the path resolves to a flow, not by job kind: excluding previews wholesale also dropped the flow editor's step test, whose path is real - page the per-case last-run lookup by created_before until the loaded cases are covered; one page of 200 reported older cases as never run - do not record an experiment when nothing launched - only attach the case input to a job when a scorer will read it - keep the case table through a save; only a different dataset clears it - drop the superseded duplicate comment on the score parser Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: stop refetching run history on every case write Reading the case list before the first await made the whole job-history query a dependency of it, so every save, delete and Load more refetched up to 1000 job rows and blanked the column. Read untracked instead. - an empty Last run cell now distinguishes never-ran from not-found-within the page bound, which the comment already claimed and the cell did not - reloading a dataset no longer replaces a populated table with a skeleton - keep the score-parser comment that describes every shape it handles Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * refactor: keep eval datasets in Postgres instead of object storage Datasets, cases and experiments become rows (`eval_dataset`, `eval_case`, `eval_experiment`, `eval_experiment_case`) rather than objects under a `wmill_eval_datasets/` prefix. What a run produced is still the job's: only case inputs and an experiment's case snapshot are stored. This removes the machinery the object store needed: - The advisory lock and the read-modify-write of a per-dataset JSONL. A case is a row, so there is nothing to serialize. - The launch-time identity check on the dataset. The foreign key makes a concurrent delete fail the transaction instead. - The trust guard on an experiment's job ids, which existed because a script can write workspace object storage directly and could forge an experiment naming somebody else's job. An experiment now chooses every job id and records itself before pushing anything, so a launch that dies partway leaves a recorded case whose job is missing rather than a running job nothing accounts for; cases that never reached the queue are removed again. Row-level security on `eval_dataset` is the authority on who may read or write a dataset, so `extra_perms` grants work and the rule is not mirrored in Rust. Cases and experiments carry a read policy derived from their dataset and no write policy: they are written on the unrestricted pool after the dataset row itself has been asked, with `SELECT ... FOR UPDATE`, whether the caller may write it. Cases are capped at 256 KiB each and 10 000 per dataset, refused rather than truncated. Attachments are S3 references, not inline bytes, so a case that approaches either cap is a mistake rather than a use case. Evals no longer need the `parquet` feature or a configured workspace object storage. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * style: align the eval drawer with the design system - Scorer chips are `Badge`s rather than a hand-rolled bordered span, and the section header is a `Label` with its tooltip, as are the case editor's fields (which also gets the label colour right). - The results table showed status as a coloured bullet, which says nothing to a colour-blind reader. It now carries the same icons the runs table uses, with the status as its accessible name. - Feedback colours move to the `-500` shades the brand guidelines name. - The conversation JSON error uses `TextInput`'s `error` prop for the border and the caption style for the message, as elsewhere. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: author an expected answer, tags and attachments on a case Every scorer is handed `(input, output, expected)`, but nothing could produce an `expected` except a conversation capture: the case editor had no field for it and a captured run left it empty. So: - The editor gains Expected, Tags and a read-only list of the attachments a captured case carries. Expected is plain text, or JSON when the answer has structure. - Capturing from an AI agent run keeps what that run answered, which is the only moment a reference answer exists for free. The results table also laid itself out by content, so a long answer pushed the scores — the numbers the table exists for — off the edge of the pane. It is fixed-layout now, with the text columns bounded. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * docs: expected is captured from a run and can be authored Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: link a saved agent when inserting an ai agent step "AI Agent" in the step picker was a leaf that always created a blank step, so reusing a saved agent meant inserting a blank one, opening its step input and linking it there. It is a category now, like Flow and AI Sandbox, listing the workspace's `ai_agent` resources next to a blank option, filtered by the picker's own search. A picked agent produces a step that is already linked rather than one linked afterwards: `agent` set, no tools, and only the flow-local `user_message`/`user_attachments` transforms. Seeding the brain keys there would leave transforms a linked step never reads and that `AgentResourceBar` strips on its next link change. Each `on:new` forwarder rebuilds the insert detail field by field instead of spreading it, so a new field is dropped unless the forwarder names it. `agentPath` is typed on both `GraphEventHandlers.insert` and `FlowGraphV2`'s `onInsert` so the next one to forget it fails the check. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: restore the link on cancel and simplify the agent bar Cancel on an agent edit forked the step into a standalone copy, which is the opposite of what the word means and needed a paragraph under the card to explain. It discards the edits and re-links the step now, leaving the agent untouched; diverging from an agent is Unlink's job, on the linked card. This flow's `tool_inputs` survive the round trip as overrides, so Cancel no longer folds them into the tools the way Unlink does. Linking a step to a saved agent happens in the step picker at insert time, so the bar's own resource picker is gone and "Save as agent" is the one action left. Its `+` button was a trap besides: it opened the generic resource form, where an agent would have to be written as raw JSON. The card itself was `surface-secondary`, the sections token, so in dark mode it was darker than the pane and read as a sunken well rather than an elevated card. It uses `surface-tertiary` as the brand table prescribes, its tool chips are `Badge`s, and the editing card no longer overflows the pane and clips its own buttons. The remaining tooltip follows the inline `Label` convention rather than sitting in a flex row whose gap stacked on the trigger's own margin. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: rework the AI agent evals surface into one table Evals become a single pane: a dataset of cases, one column per scorer, one row per case, with the run being looked at chosen from the toolbar. Runs are permanent. Running the whole dataset opens one; running a single case records nothing at all — it is a job, and looking at what it did is not a claim that it belongs in the history. Its result and its scores sit over the row until they are saved as a run, which carries the cases that were not rerun and the scoring jobs themselves, so the number that is saved is the number that was looked at. A scorer is a runnable: a judge agent or a script, created in one click and edited in place. Scores carry a reason and per-assertion checks, shown on hover with a rescore button. What ran is always named. A run records the agent version, or — for a configuration that is not deployed — a hash of it, so a table can say that its numbers describe an agent that no longer exists: those rows dim and the table offers to rerun. An agent's draft can be run directly instead of the deployed value, and once those edits are deployed the runs that made them are recognised as that version. A step with no agent of its own is evaluable too, and saving it as an agent moves its history onto it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: keep an agent's in-progress edits on the agent Editing a linked agent forks it into the step, which is what makes the edits runnable there — but the agent is what is being edited, so that is where the unsaved state belongs. The edit is mirrored into the agent's own resource draft as it is made. It then survives leaving the flow, shows the agent as drafted wherever it appears, and is what evals run when asked to run the draft rather than what is deployed. Deploying or cancelling clears it; opening Edit without changing anything does not mark the agent as drafted. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: shape the evals surface around a saved agent Evals hang off an `ai_agent` resource, so the surface is now only ever about one: the `draft` subject kind, the standalone-step subject and the move that carried a step's history onto a newly saved agent are gone. - A run is permanent and numbered per agent. Running a single case is a trial: it answers in the panel and never touches the table. - "Run scorers only" opens a run of its own that reuses the answers of the run you are looking at, so a scorer added later measures what already ran without calling the agent again. - A draft run whose configuration is later deployed is stamped, once, to the version it became, so its label stops reading `v23 + edits` forever. - A scorer can carry a pass threshold, read off the scores already recorded. - The table is the case, its answer and one number per scorer; datasets are created and edited in a drawer; a run that executed an earlier state of the current draft says so above the table, in one line. - Which agent a step is, whether it is being edited, and which version it is on is a strip above the step's tabs, because it is true of every tab. - Capturing a case from a step test or a conversation is dropped, and with it the `memory` override on a linked step that nothing set. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: run past versions of an agent, and number versions per resource The evals home becomes one table of every run of the agent, whichever dataset each is of, with one badge per scorer. A list spanning datasets cannot hold every dataset's scorers to look a name up, so a score carries its name and kind with its number, and thresholds are joined in per run and column. Run now asks what to run: the latest agent, resolved when the run executes as a flow step does, any past version, or the unsaved edits. Pinning is a subject kind of its own, since a linked step resolves the resource live and inlining is the only way to run a version that is no longer current. Scorers move into the edit-dataset drawer. The column header over a run reports and nothing else: a run is permanent, and a control there that changed the columns would edit the past from the one place that must not. Adding one offers four ways rather than two, writing and reusing being different jobs, and both new kinds open with a summary filled in. Versions are numbered per resource. `resource_version.id` is one identity sequence for the whole table, so an agent saved nine times read v4 ... v24, and the gaps counted writes in workspaces the reader cannot see. The id stays how a version is addressed; the new number is what it is called, in the resource history drawer as well as here. It is assigned on write rather than counted on read because trimming past the cap and clearing a history both take the oldest rows, and counting the survivors would renumber a version a run already names. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: read the dataset a remembered selection names Reopening the evals modal restored the last dataset from storage as a bare path, without reading the row it names. Every "is this already the one?" test compared against that selection, so all of them short-circuited and the dataset was never loaded: editing it opened a drawer with no summary, no scorers and no cases. The remembered path is now brought into context the same way any other choice is, and the tests compare against the dataset that is loaded rather than the one that is selected, so a selection can no longer stand for a read that did not happen. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: give dialogs a trail in their header A dialog deep enough to navigate had nowhere to say where you were: the header held a fixed title, and the way back was a control each body placed for itself, somewhere in a toolbar that moves with everything else the toolbar holds. The header is the one part of the surface that does not move, which is where the trail belongs. `Modal` takes an optional `trail` of levels below its title, rendered as a breadcrumb whose ancestors are the way back. Declarative on purpose: callers of this depth already hold the state that says where they are, so the dialog reads it rather than owning a stack they would have to push and pop in step with it. Escape follows the trail. Leaving a level is what someone deep in a dialog means by it, and closing the whole surface throws away the navigating they did to get there; at the root it closes as before. That only works if a dialog can tell it is the surface being addressed, so `Disposable` now answers `isTopmost()` and the dialog asks before acting: it keeps Escape for itself, so nothing else was arbitrating between it and a drawer opened from inside it, and both were acting on one key press. Evals is the first caller: its runs list is the root, a run is a level in it, and the back button that used to sit above the table is gone. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: portal dialogs out of wherever they were opened from A dialog rendered in place inherits whatever the calling component happens to sit inside. One `transform`, `filter` or `overflow` anywhere above it makes its `fixed` positioning resolve against that ancestor instead of the viewport, and a surface meant to cover the app is then confined to a box it never asked for: the nav rail paints over it and its own edges are clipped. Drawers have always portalled for this reason. Dialogs only did so when an enclosing pane claimed them, and rendered in place otherwise, so the same screen could show a drawer over everything and a dialog trapped behind the nav. They now portal the same way: to the pane when one claims it, to `body` otherwise. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: make the dialog's title the first step of its trail The trail listed levels below the title, so a dialog one level deep read "Evals > All runs > Run 20 · v6": three steps for two places, the first two of them the same place under different names. The title is the root, so it is the root's own segment, and the trail a dialog is given is now the whole path with that segment at its head. Its height stopped moving too. A heading carries a line-height of its own, so a header holding only an h3 stood six pixels shorter than one holding segments as well, and the dialog's whole top edge stepped as you navigated. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: sharpen the evals controls around where you are standing Each screen now offers what belongs to it. The list starts runs; a run is a record, so it offers only the one thing that acts on the record itself, which is measuring the answers it already stored. Starting a fresh run from inside one asked which agent and which dataset from the screen least about either, and scoring an existing run was offered from the list, where there is no run to score. Which run and what it is read against are one question asked twice, so they sit together rather than at opposite ends of a row. Choosing what to run is now a toggle over the two states worth naming, the draft and the saved agent, with every earlier version one click further: running an old version is deliberate, and a list made all three look alike. The draft is read when the dialog opens rather than taken from the caller's polled copy, which could be seconds behind an agent edited a moment ago and would leave the option out exactly when it is the reason for opening the dialog. The dataset field carries its path under it and its edit button on hover, as a resource picker does, so the closed field says what the open list said. Edits waiting on an agent are a "draft" here as everywhere else in Windmill, rather than "+ edits". The dialog runs an evaluation rather than "the agent", which is what it was already called everywhere it is recorded. An agent being edited keeps its evals button on a line of its own, clear of the decision to save or discard. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: settle the evals controls on the patterns Windmill already has The version choice uses ToggleButtonMore, as the AI provider picker does: the two states worth naming stay in the group, the rest are behind the overflow menu, and the one you pick joins the group rather than appearing in a second control below it. The deployed one says which version it resolves to. A run offers nothing to start. Scoring an existing run again was the last thing left there, and it was one button explaining a distinction that the run and the dataset already make between them. The warning that a run executed an earlier draft is about the run on screen, so it goes when the run does rather than following you back to the list, and it sits against the table instead of inside a frame of its own. A dataset just created stays open for its scorers and cases: those are what a dataset is, they can only be added to one that exists, and closing on create sent you to find it again to add them. Scorer settings are a cog rather than a word, now that the row holds three actions. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: close the gap in the version toggle and say what naming a dataset does The overflow trigger is not a pill, so the room it reserves showed as a gap between it and the button before it; it is pulled in by that much. The dataset field gets its clear button, which is also the slot the edit button is positioned against, so the two now sit where a resource picker puts them. Naming a new dataset said nothing about what happens next, and the drawer looked like it was missing the rest of itself. It says so instead: a scorer and a case both belong to a dataset, so there is nothing to attach either to until this one exists, and creating it leaves the drawer open on them. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: choose a dataset's scorers while naming it A scorer is a reference to a runnable, not a child of the dataset, so it needs the dataset's name but not its row. The list is collected in the drawer while the dataset is being named and sent with the create, which already accepts one, so a dataset arrives holding the columns that were chosen for it rather than being made empty and then edited to hold them. Cases stay where they were: a case *is* a row of the dataset, so there is nothing for it to be a row of until one exists. The drawer says which of the two is which instead of leaving the screen looking like it is missing the rest of itself. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: level the version toggle and name the dataset in its own field The overflow trigger stands a row taller than a toggle button, so the group grew to its height and left the sunken background showing under every pill beside it. Every child of the group is the same height now, which is why the AI provider picker never had the band: it sizes them all alike. The dataset field says the summary with the path after it rather than carrying the path on a line below. The list stacks the two, which a one-line field cannot do, so it says both the other way round. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: tidy the evals forms and the run's own controls Picking a scorer that exists chooses between two sources rather than showing both: the ones already measuring something, and everything else in the workspace. The first list says what each is called with its path under it and what it already measures on the right, instead of three columns that were the same path truncated three ways whenever a scorer had no name of its own. A dataset's drawer says what it is for on the page rather than under an icon, and its summary is sized like the field beneath it. The run's own row lines up with the table under it, the warning above that table is spaced off the rule rather than sitting on it, and adding a case is gone from a run: a run is a record of cases that were answered, so curating them from it is editing what it measured. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: create a dataset holding the cases written for it Creating a dataset takes the cases to create it with, so one can be assembled in a single act instead of made empty and then filled in. The drawer holds them while the dataset is being named, gives them ids of its own to be edited by, and sends them with the create. Every case is checked before the dataset is written. `eval_case` grants users no write, so the rows cannot be inserted in the transaction that creates the dataset under the caller's own policies; validating first is what keeps "created holding these cases" from becoming "created, holding some of them", and the rows that do follow go in one transaction of their own. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: name the button for what it opens, and say what each version is Starting an evaluation asks which state of the agent and which dataset, and both cost a provider bill, so a button that read as spending one on the way past was lying about the click. It opens something, and says so. Running one case from the panel keeps its own name and its play icon, because that one does run on click. The version options say what they are rather than what they are not: what a flow step would or would not run is a fact about somewhere else, and someone choosing what to evaluate is not standing in it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: give the editing card two rows and mark evals as beta At the width of a step panel the card's one row wrapped: the line naming the agent, the line saying what saving does, and the two buttons deciding the edits' fate all fought for it. Deciding gets a row of its own, and evals sits against the line it is about, since evals of an agent being edited run the edits. Evals is named wherever it is offered. It read as a word in one state of the card and as an icon in the other, which is two things to recognise for one door. The dialog carries a beta badge against its own name, before any level below it: every way in lands there, so it is said once and stays put as you navigate. The version toggle spells out which is which. Both are the agent at v2 and the difference between them is the whole choice, so it is worth the width. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: name a new dataset, and lay the scorer's settings out like a step's inputs A new dataset arrives called "Dataset 1", which the path follows as it follows any summary: a dataset with none was one every table could only call by its path, and the two seeds are what the summary rule already produces. Scorer settings put each field's description between its label and its input, where a step's inputs put theirs, and its inputs are the size the rest of the drawer uses. The runnable behind the column is a link to it with its kind's icon, since it is a resource of its own and the one thing about it these fields cannot change. The line explaining that a pass line re-reads recorded scores went: the threshold is a number to set, and how it is applied is not a decision being made here. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: curate a dataset in the drawer and save it in one act The drawer holds the cases while they are edited and writes them when it is saved: added, changed and dropped, whichever it is. Typing no longer writes, so a set is never half saved while someone is still deciding what is in it, and Save means the same thing whether the dataset exists yet or not. A case panel offers reading rather than acting. Running one case now and editing one from a run were the last two ways to change a record from the screen showing it, and the machinery behind the first went with it. The answer is rendered as the prose it is, under what it is: the case's result, whichever run is selected above it. The rest is what the run's table was doing to its own edges: a column name is clipped to its column rather than running into the next, the table squares off against an open panel, and that panel closes with the run it belonged to. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: one border above a table, and a link to the run's job The row above the table drew a bottom border and the table draws its own top edge, so every table sat under two lines. The row keeps its spacing and the table keeps its edge. A column header no longer spins while its scores arrive: the cells under it are where the numbers are missing, and they say so themselves. The beta badge is the height of the word beside it rather than of the line it sits on. A run is one flow and therefore one job, so the run says where that job is: what it is doing, what it cost and what it logged are all there rather than reconstructed from the table. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: stream scores as each scorer finishes, and show them per case A scorer runs after the agent inside the case's own iteration, so its verdict can be read as soon as its step is done. Waiting for the iteration to end held every column of a case back until the last of them finished, which is why answers arrived one at a time and scores all at once. Reading a job that is still running needs one guard: a module with nothing in it is a step that has not run, not one that produced nothing, and recording the second makes a failure that never goes away. The panel beside the table shows what each column made of the case and why. The reason a judge gave was stored and never shown, which is the half of a score that says anything. It stops repeating the question the header already asks, and a case still running reads as waiting rather than as an answer that says "Running". A run is a number beside a dataset, so the list puts the two together. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: score a case with every scorer at once The scorers of a case read the answer and never each other, so they ran one after another for no reason: measuring a case now takes as long as its slowest column rather than as long as all of them. Each is a branch of its own, kept from failing the others, so a judge that errors costs its own column and no more. An iteration is three steps again — answer, payload, scores — rather than one per scorer, and each branch is named for the column it produces, so the graph of a run says which scorer did what instead of spelling out an id. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: read a judge's score out of the JSON it nearly wrote A judge quoting the agent inside its own reason writes those quotes unescaped, which is invalid JSON and also the most ordinary sentence for it to produce. The whole verdict was being thrown away over it, so a column that had a number reported having none. The number and the reason are now read straight out of such text. Deliberately not a second JSON parser: it finds the two keys and takes what follows, which is what survives a quote in the middle of a sentence. A case still running says so with a spinner rather than with the word "Running" sitting where its answer goes. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: ask a judge for a shape instead of trusting it to write one A new judge carries an output schema, so the provider holds it to `{score, reason}` rather than the prompt asking it to. Windmill already delivers a schema whichever way the model takes it, a tool for Claude and Bedrock and the native parameter elsewhere, so there is no list of models to keep here. An agent with no runs offers its first one where the first row would be, rather than from a toolbar above a table that has nothing in it. Starting a run no longer picks a dataset for you. It fell back to whichever came first, which on an agent that has never run means offering another agent's set as though it were the obvious one; and with no dataset at all it says so and offers the one move there is. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: report a column that failed throughout, and hold the run dialog The runs overview dropped any column that produced no number, so a judge that failed on every case of a run vanished from the row and read as a column nobody had asked for. The aggregate now reports every column that has cells, with the count of the ones it failed on, and the badge says "failed" where there is nothing to average. A column with no cells at all is still left out: that one was added after the run and has nothing to say about it. Creating a dataset closes the drawer rather than turning it into an edit of what it just made: scorers and cases already ship with the create, so there is nothing left to stay open for. Reached from the run dialog, it gives the screen back with the new dataset selected, and the dialog keeps the version you had already chosen. Also: - the case panel's job link moves to the panel's own header, where its scope is: the job is the whole iteration, not the answer it sat over - one action in the scorer drawer's header, as its neighbours have. The reuse list picks rather than adds, and says which dataset each column already measures - adding a case is the last row of the list it lands in - the pane shows what it has read rather than an empty state it has not earned yet, and its rows say they open - the linked agent card loses a border it had inside another one * fix: keep the linked agent card's outline The card is a thing inside the step's inputs rather than a section of them, and the outline is what says so. Only the rule inside it goes: the detail it separates is already set apart by being detail. * refactor: fit the eval surface to the shipped design * feat: give a nested dialog a back control and the runs list its own moves Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat: put a dialog's description under its title Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * refactor: fold a dialog's back control into the crumb it returns to Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat: edit a dataset's cases as a table rather than a list beside a form Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat: edit a dataset's cases in the grid the data tables are edited in Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat: edit a grid cell of prose in place, and cap a dataset at one page Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * refactor: keep the cell editor's styles beside it, not in the vendored theme Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: keep an empty cell empty and cap the editor's growth Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat: name the step that assembles a run for the scorers Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat: run the payload step natively, and say so when nothing serves that tag Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: report an answer as answered while its scorers are still running Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat: let a scorer say a case is not one it measures Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: score the answer, and leave a case with no expected answer unmeasured Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * refactor: split the evals backend into modules Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: record what a run produced so it outlives its jobs * fix: read only the agent step's own tool jobs into the payload * fix: pin a run's configuration and give the judge the attachments * feat: write a dataset's cases in one transaction * chore: refresh the sqlx cache for the eval queries * fix: drop results a newer selection has superseded * fix: keep a draft the agent editor never opened on * feat: let a run record what it produced instead of waiting to be read * fix: serialize the replacements of a dataset's cases * fix: stop the poller from superseding a read slower than its interval * chore: refresh the sqlx cache * fix: keep a failed read from settling a cell as a case with no answer * fix: hold the case grid while its save is in flight * fix: keep a failed collect step from failing the run it recorded * chore: refresh the sqlx cache * fix: commit an open cell into the save that reads it * refactor: size the eval buttons with unifiedSize * docs: describe a run as the one flow it is * fix: show a run's recorded rows when part of it cannot be collected * refactor: size the remaining PR-added buttons with unifiedSize * fix: save the dataset name that was submitted, not the one typed after * fix: force an open cell into the save that was pressed for it * fix: refuse to score a run whose evidence could not be read * fix: hold one lock over a dataset's case count and its writes * fix: keep one unreadable run from costing the whole runs list * refactor: drop the banned bindable-default from the eval props * fix: hold the scorer controls while the dataset is written * fix: read only the caller's own draft of an agent * docs: say in the contract that a run pins its configuration * fix: say a scorer did not run rather than blaming a missing answer * feat: resume the agent draft you already had when you press Edit * refactor: build the trail and dataset controls from Button * fix: clear the open-cell flag when the drawer reopens * chore: refresh the sqlx cache * fix: read a run's configuration and its version from one snapshot * fix: refuse a dataset path or summary the column cannot hold * refactor: handle the agent draft the way the resource editor does * fix: run only a configuration the launch actually read * docs: bound dataset path and summary where they are submitted * fix: surface a stalled agent draft instead of claiming it is kept * fix: stop claiming a draft holds edits a failed write never sent * fix: word a missing score only once the run says whether the case answered * fix: let a breadcrumb crumb shrink so its truncation applies * docs: describe where an agent's unsaved edits live and what drops them * fix: keep harvesting scores when the run cannot yet word a missing one * fix: report a refused draft write the card was reading as a save * fix: drop the refused draft write when the server copy is taken instead * refactor: build the scorer and dataset pickers from the design system * fix: say what removing a scorer column actually does * fix: drop a refused draft write wherever the server copy is read * fix: let a picker row be as tall as the two lines it holds * docs: record what removing a scorer column does to recorded runs * fix: send a queued draft write before reopening, and drop only what it refuses * refactor: write the agent draft at commit points instead of mirroring keystrokes Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * refactor: run an agent's edits from the step instead of keeping them as a draft Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: make the diff badge keyboard operable and refuse an edits run without its edits Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: drop the dataset icon from the scorer picker rows Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: size the evals buttons like the rest of windmill and call a run of edits edits Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: count a brain expression as an edit of the linked agent * fix: cap scorers per dataset and report a launched run as launched * fix: harvest scores in one read, refuse duplicate case ids, allow group paths * fix: mint scorer ids server-side, save a dataset edit in one request, check attachments * fix: write a dataset edit and its cases in one transaction * fix: atomic dataset create/edit, reset eval pane per agent, stable pending scorer ids * refactor: govern eval_case writes by RLS so a dataset edit is one transaction * fix: pin launch snapshot, order case locks, cap dataset size, guard stale load * fix: cap dataset bytes on single-case writes, reset run-dialog flag on load failure * feat: migrate eval datasets on username change, settle unspawned cases, drop unused case endpoints * fix: resolve scorer scripts as the caller and pin their hash; migrate scorer paths on rename * fix: bound a failed tool call's error to the payload truncation cap * fix: pin scorer hash as a hex string, reject missing judges, migrate eval authorship * fix: record an out-of-range scorer result as an error, not a score * fix: resolve judges in one caller-scoped read, pin deployed scripts, bound pass_if * fix: settle unspawned cases only when the run completes, and their score cells too * feat: reassign eval datasets and their path references when offboarding a user * fix: use the regex backreference in offboarding eval path rewrites * fix: register eval datasets in offboarding registries, keep resource-version param name * refactor: name the resource-version path param id, since it is the row id not the version * fix: validate dataset paths canonically, clone eval data on fork, surface eval load and launch failures * docs: note MCP tool results are not yet surfaced to eval scorers * fix: show the eval error state on any load failure, not only an empty dataset list * fix: preserve eval case order across a batched save Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * docs: scope the eval launch delete-safety guarantee to the assembly window Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * fix: only offer deployed scripts as eval scorers, drop unbuilt rescore claim Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * fix: enforce 0-1 scorer threshold in the settings drawer and clear stale eval load errors Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * fix: scope subject version/hash reads to the caller and keep a 0 pass threshold Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * fix: select the saved dataset when creating or renaming from the Run dialog Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * fix: gate eval dataset rename on path ownership, not just write access Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * fix: tolerate a malformed agent config when resolving the deployed label Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LVjewUvXFEjNLLw7kxz41h * refactor: trim eval code and comments, fix shared select and modal paths Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: drop the rename warning when editing an eval dataset path Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: add eval dataset delete, keep summary on partial edits, settle resultless scorer cells Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * test: cover parseThreshold and subjectLabel Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix: hold dataset Save during a scorer write, derive draft_hash only from the carried draft Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
b6e059116a |
feat: track token cost in AI sessions and chats (#10688)
* feat: track token cost in AI sessions and chats * fix: address review findings on AI cost tracking * fix: price inherited and overridden models at their real rates * fix: stop newer model revisions inheriting an older price * fix: stop a sub-model inheriting its family's price * fix: keep alias suffixes resolving to their model's price * fix: count OpenRouter cache writes and drop unverifiable rates * refactor: move AI spend out of the chat into workspace and user settings * fix: pin the usage workspace per turn and stop inventing cache rates * fix: leave Sonnet 5 unpriced while its promotional rate runs * docs: record the new table in the schema summary and tighten comments * fix: mark estimated AI costs with ~ and drop session grouping Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: name the workspace in the self-scoped AI usage title Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * docs: state that overrides never replace a provider-returned cost Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: let a cleared cache rate inherit again and flag partial totals Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: clear a refused rate's error when the input snaps back Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: stop a revision variant inheriting its base family's rate Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: report AI usage before tools run and price self usage consistently Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: key pricing rows on the model id usage is reported under Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: surface Bedrock and Gemini usage the chat proxy was dropping Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: count Gemini tool-use prompt tokens as input Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: price flat-rate Gemini Flash models Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * docs: state the tool-use token invariant once Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Diego Imbert
|
7751d3e43e |
feat(frontend): warn when COEP blocks cross-origin resources in raw app editor preview (#10328)
* feat(frontend): warn when COEP blocks cross-origin resources in raw app editor preview Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Gr3udvsSKYyH6mnDvGDqEE * fix(frontend): hedge COEP toast wording and attach warning on detached preview initial load Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Gr3udvsSKYyH6mnDvGDqEE * chore(frontend): condense COEP warning rationale comment Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Gr3udvsSKYyH6mnDvGDqEE --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
74af4ed939 |
chore(main): release 1.795.0 (#10807)
* chore(main): release 1.795.0 * Apply automatic changes --------- Co-authored-by: rubenfiszel <275584+rubenfiszel@users.noreply.github.com>v1.795.0 |
||
|
hugocasa
|
dc27db68de |
fix: require item read scope on workspace tarball export (#10797)
* fix: require item read scope on workspace tarball export Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: accept a wildcard path grant for whole-domain scope checks Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: let a wildcard path grant delegate the unqualified scope Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> Co-authored-by: Ruben Fiszel <ruben@windmill.dev> |
||
|
hugocasa
|
40f0cab2ad |
fix: scope capture deletion to the workspace in the request path (#10795)
* fix: scope capture deletion to the workspace in the request path Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test: layer the capture fixture on base instead of duplicating it Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
5b885ae311 |
fix: keep raw-app files within their app folder on sync pull (#10796)
* fix: keep raw-app files within their app folder on sync pull Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: validate raw-app file keys as stored, closing nul and duplicate-field bypasses Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: guard raw-app runnable ids too and fail closed on unparseable value Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: strip only a leading slash on raw-app file keys to match backend Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: strip only a leading slash on raw-app file keys to match backend Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> Co-authored-by: Ruben Fiszel <ruben@windmill.dev> |
||
|
hugocasa
|
25d9a20630 |
fix: require an unscoped token to reach the workspace encryption key (#10798)
* fix: require an unscoped token to read the workspace encryption key Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: hold the encryption key's write path to the same token bar Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: audit a workspace export only once nothing can still reject it Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore: carry the new audit operation into the served openapi spec Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
01fc4f1568 |
fix: name the requested storage when a workspace storage lookup finds nothing (#10803)
* fix: name the requested storage when a workspace storage lookup finds nothing * chore: point ee-repo-ref at the merged ee commit |
||
|
Ruben Fiszel
|
a350f7c68e |
feat: show the date on the runs dashboard chart axes (#10808)
* test: assert the unpacked repo symlink without following it `unpack_keeps_a_link_that_stays_in_the_repo` read through the link it had just unpacked. Windows stores a symlink's target verbatim and its object manager rejects the `/` in a POSIX one, so `read_to_string` came back with `ERROR_INVALID_NAME` and the release's `cargo_test_windows` job was red. Pin what the function is responsible for on every platform — the link is kept and materialized — and read through it only where a POSIX relative target resolves. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * test: key the cli sync-map fixtures with the platform separator A sync map is keyed with the platform separator on both sides — `FSFSElement` walks the tree with `path.join`, and the remote `ZipFSElement` starts at `"." + SEP` and joins from there — while an `!inline` reference is always forward-slash. `lock_dedup.ts` follows that convention; the fixtures did not, so on Windows they built a map shape the CLI never produces and 12 of them failed. `getTypeStrFromPath` is the same story: it matches `"dependencies" + SEP`, and the test handed it a forward-slashed path. Build the fixture keys through the separator, leaving the `!inline` references and the `present` map forward-slash, as `sync.ts` hands them over. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * ci: skip the discord comment relay when the thread lookup returns none A rate-limited or unauthorized Discord response carries no thread list, and under `bash -e` that aborted the step — jq cannot iterate null, nor parse the HTML error page Cloudflare answers a 429 with — before it reached the "thread not found, skipping" branch right below. Three comment relays failed that way on the 1.794.0 head. Keep the step green for both, but tell them apart: a response with no thread list is a delivery that was dropped for a reason worth seeing, so it warns with the body it got, while a PR that genuinely has no thread stays quiet. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * feat: show the date on the runs dashboard chart axes Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019UA8Kikr1QD28g2fyWoSbj * fix: keep the runs chart date visible on sub-day ranges Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019UA8Kikr1QD28g2fyWoSbj --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
4b406e37c0 |
fix: size the ephemeral job token to the job timeout it must serve (#10804)
* fix: size job token to the premium cloud job timeout Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: give the job token setup headroom and drop dead MAX_TIMEOUT_DURATION Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: cap job token setup slack so self-hosted tokens stay at 7d Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
01891cd732 |
fix: keep workflow-as-code scripts off dedicated workers (#10805)
* fix: keep workflow-as-code scripts off dedicated workers A dedicated subprocess calls the script's `main`. A workflow-as-code v2 entrypoint exports none, so a WAC script configured as a dedicated worker failed every run with `entry.module.main is not a function`, and its checkpoint/dispatch round-trip never ran at all. Leave such a script unregistered in the dedicated worker map instead. The worker still holds the script's dedicated tag, so the job falls through to the regular executor on the same worker and runs correctly; rejecting it at push time would strand it, since nothing else pulls that tag. `is_wac_v2` covers only the languages whose executor actually routes a workflow through the WAC runner: Deno runs a WAC-shaped script as a plain `main`, so claiming it is WAC would deny it a path it uses correctly today. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01VMNiDoBuUY9jzqcuFLT2wU * chore: update ee-repo-ref to ac02c4696ea0be6a8b8ae154ddd7521bc1c3bbc0 This commit updates the EE repository reference after PR #740 was merged in windmill-ee-private. Previous ee-repo-ref: bf742f6ea4d435bd47c9ee0ac5ad800925d79672 New ee-repo-ref: ac02c4696ea0be6a8b8ae154ddd7521bc1c3bbc0 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> Co-authored-by: windmill-internal-app[bot] <windmill-internal-app[bot]@users.noreply.github.com> |
||
|
Ruben Fiszel
|
1a506b8f22 |
chore(main): release 1.794.1 (#10801)
* chore(main): release 1.794.1 * Apply automatic changes --------- Co-authored-by: rubenfiszel <275584+rubenfiszel@users.noreply.github.com>v1.794.1 |
||
|
hugocasa
|
e0510fea21 |
fix: keep every value of a repeated multipart field (#10800)
* fix: keep every value of a repeated multipart field Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore: drop pre-change narration from a multipart test comment Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
dc0df45c81 |
chore: bump git-sync hub scripts to windmill-cli 1.794.0 (#10802)
* test: assert the unpacked repo symlink without following it `unpack_keeps_a_link_that_stays_in_the_repo` read through the link it had just unpacked. Windows stores a symlink's target verbatim and its object manager rejects the `/` in a POSIX one, so `read_to_string` came back with `ERROR_INVALID_NAME` and the release's `cargo_test_windows` job was red. Pin what the function is responsible for on every platform — the link is kept and materialized — and read through it only where a POSIX relative target resolves. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * test: key the cli sync-map fixtures with the platform separator A sync map is keyed with the platform separator on both sides — `FSFSElement` walks the tree with `path.join`, and the remote `ZipFSElement` starts at `"." + SEP` and joins from there — while an `!inline` reference is always forward-slash. `lock_dedup.ts` follows that convention; the fixtures did not, so on Windows they built a map shape the CLI never produces and 12 of them failed. `getTypeStrFromPath` is the same story: it matches `"dependencies" + SEP`, and the test handed it a forward-slashed path. Build the fixture keys through the separator, leaving the `!inline` references and the `present` map forward-slash, as `sync.ts` hands them over. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * ci: skip the discord comment relay when the thread lookup returns none A rate-limited or unauthorized Discord response carries no thread list, and under `bash -e` that aborted the step — jq cannot iterate null, nor parse the HTML error page Cloudflare answers a 429 with — before it reached the "thread not found, skipping" branch right below. Three comment relays failed that way on the 1.794.0 head. Keep the step green for both, but tell them apart: a response with no thread list is a delivery that was dropped for a reason worth seeing, so it warns with the body it got, while a PR that genuinely has no thread stays quiet. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * chore: bump git-sync hub scripts to windmill-cli 1.794.0 --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
5088e13705 |
fix(ci): unbreak the windows test jobs and the discord comment relay (#10799)
* test: assert the unpacked repo symlink without following it `unpack_keeps_a_link_that_stays_in_the_repo` read through the link it had just unpacked. Windows stores a symlink's target verbatim and its object manager rejects the `/` in a POSIX one, so `read_to_string` came back with `ERROR_INVALID_NAME` and the release's `cargo_test_windows` job was red. Pin what the function is responsible for on every platform — the link is kept and materialized — and read through it only where a POSIX relative target resolves. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * test: key the cli sync-map fixtures with the platform separator A sync map is keyed with the platform separator on both sides — `FSFSElement` walks the tree with `path.join`, and the remote `ZipFSElement` starts at `"." + SEP` and joins from there — while an `!inline` reference is always forward-slash. `lock_dedup.ts` follows that convention; the fixtures did not, so on Windows they built a map shape the CLI never produces and 12 of them failed. `getTypeStrFromPath` is the same story: it matches `"dependencies" + SEP`, and the test handed it a forward-slashed path. Build the fixture keys through the separator, leaving the `!inline` references and the `present` map forward-slash, as `sync.ts` hands them over. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx * ci: skip the discord comment relay when the thread lookup returns none A rate-limited or unauthorized Discord response carries no thread list, and under `bash -e` that aborted the step — jq cannot iterate null, nor parse the HTML error page Cloudflare answers a 429 with — before it reached the "thread not found, skipping" branch right below. Three comment relays failed that way on the 1.794.0 head. Keep the step green for both, but tell them apart: a response with no thread list is a delivery that was dropped for a reason worth seeing, so it warns with the body it got, while a PR that genuinely has no thread stays quiet. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01P3WRtxdKNGdomWX9vaAYGx --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
97ff901176 |
fix: parse the hub resource type schema instead of relaying it as a string
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
49a4743089 |
test: report metadata_source from whether the fixture has authored meta
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
a4279215a1 |
test: cover two more real integrations where the scripts already answer
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
hugocasa
|
850f0d46ec |
test: benchmark the hub tool against a real integration's scripts and metadata
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
55b6279058 |
chore(main): release 1.794.0 (#10782)
* chore(main): release 1.794.0 * Apply automatic changes --------- Co-authored-by: rubenfiszel <275584+rubenfiszel@users.noreply.github.com>v1.794.0 |
||
|
AlexRV12
|
0b3dc3e5c9 |
fix: build the global chat's prompt identity from the operating workspace (#10793)
* fix: do not read an unloaded workspace list as a non-membership `roleForWorkspace` settled `not_a_member` from `userWorkspaces` alone. That store and `superadmin` both start undefined and load asynchronously, so an unloaded list read as an empty one: a chat operating on any workspace other than the one being browsed advertised no pages and reported an access denial. The root layout gives up after its retries, so a load that fails leaves the denial permanent, with `whoami` never attempted. Settle a non-membership only once both stores have resolved; treat unresolved as unknown and fall through to the `whoami` lookup. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: build the global chat's prompt identity from the operating workspace The prompt's path conventions and folder guidance came from the ambient `userStore`, which describes the workspace being browsed rather than the one the chat operates on. Three of those fields are per-workspace and wrong whenever the two differ: the username (that workspace's `usr` row), the writable/readable folder sets (its ACLs), and `is_admin`, which decides whether the folder list reads as exhaustive. The backend still enforces the ACLs, so the cost is prompt quality — paths the model cannot write to, and a 403 to recover from. Resolve the identity for the operating workspace and feed that to the prompt, refreshed alongside skills and MCP servers and settled in `beforeSend` so the cached system-prompt prefix stays stable for the turn. An unresolved role now leaves the folder sets undefined rather than empty, so the guidance is dropped instead of claiming there is nothing to write to, and `create_folder` credits the workspace it wrote to. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: read the AI provider resource types lazily `Object.keys(AI_PROVIDERS)` at module scope made `AI_PROVIDERS` a load-time requirement for every importer of this module, the global chat included. `AIChatManager.test.ts` mocks `../lib` without it and has been unable to load since the catalog was introduced; no CI workflow runs vitest, so nothing reported it. The constant is read in two places, both inside functions, so deferring it removes the load-time dependency without changing behaviour. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
9022dc9d44 |
fix: confine job tokens to workspace-scoped API routes (#10631)
* fix: confine job tokens to workspace-scoped API routes Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep the object-storage connection test reachable from a job token Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep the workspace-exists check the CLI makes reachable from a job Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * test: reconcile the job-token caps after #10124 The workspace-confinement middleware answers a workspace-less route before the privilege gate behind it runs, so the cases #10124 added on those routes now see 403 rather than 401. Rejection is what they assert, but two of them needed more than a status change: - `list_worker_groups` asserted only that the response body omits the static env value, which an error body satisfies for the wrong reason. It now asserts the status, keeping the secret check as a second assertion. - `require_super_admin` lost its only unshadowed route. `GET /api/w/{workspace}/users/list_addable` is gated solely by that call and names a workspace, so it reaches the gate and pins it at 401. The module doc states the two-layer rule once; the file covers both caps, so it is no longer named for either one alone. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * docs: correct the workspaces/exists rationale and the parquet gate note `workspace` carries no row-level security, so `exists_workspace` running through `user_db` does not filter by membership as the comment claimed. State what the route actually discloses — whether a workspace id is taken. The object-storage case explained why a 404 would satisfy the assertion for the wrong reason, which described the earlier `assert_ne!(403)`; against the 422 it now asserts, a 404 fails. Say instead why the case is gated on the feature. * fix: let a job token keep the workspace-less routes that carry no workspace Confinement refused every route outside the allowlist, including ones that answer purely from the caller's own account or from the request body. Those cross no workspace boundary, so refusing them buys nothing: - `users/email` returns a value already inside the token, and `workspaces/allowed_domain_auto_invite` tests the caller's own address against a static list. Neither opens a transaction. - `users/usage` reads the caller's own row; `users/tutorial_progress` reads and upserts a UI bitfield keyed on the same email. - `schedules/preview` takes no `ApiAuthed` at all — it computes the occurrences of the cron expression in the body and returns nothing the caller did not send. The rule, not the list, is what the doc comment states: answers from the caller's own account, the request body, or content identical for every workspace; never naming another workspace, never instance configuration. The candidates it excludes are written down with their reasons, since `users/list_invites` reads as caller-scoped until you notice the response carries a workspace id per invite. Regression covers both directions — the new entries answer, and the rejected caller-scoped reads stay refused. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test: reach the privilege gates confinement hides `require_devops_role` and `require_instance_admin` gate only workspace-less routes, so confinement answers every request that would reach them: no HTTP case can tell whether they still cap job tokens, and both could lose that check with this suite green. `require_super_admin` has a workspace-scoped route to reach it; these two have none, so call them directly instead. The identity used is the fixture's real superadmin, so the passing half proves the rejection keys off `job_id` rather than off the user. Also separate two claims the write-allowlist doc had merged into one sentence: no entry writes outside the caller's own account, but what each may read differs, and the workspace-existence check answers for any id. * docs: say that the object-storage probe writes The write-allowlist lead claimed no entry writes state outside the caller's own account. `test_s3_bucket` puts an object into the store the body names and deletes it again, so it does write; a failure between the two leaves the object behind. The invariant that holds is about Windmill state. Say so in the lead, and describe the put/delete in the entry itself rather than leaving "acts only on the store the request body describes" to imply a read. * test: cover the last two job-token gates confinement hides Seven guards key on `ApiAuthed::job_id`. Three keep a workspace-scoped route and are exercised over HTTP; the other four are reachable only through workspace-less routes, which confinement now answers first, so nothing observed whether they still cap job tokens. `require_devops_role` and `require_instance_admin` were already called directly. Add the two that were not: `forbid_superadmin_job_token`, and `forbid_elevated_job_token`, whose call sites are `create_token`, `update_token_scopes` and `set_password` — all workspace-less. Both key on two conditions rather than one, so all three combinations are pinned: neither fires without job provenance, and neither fires for an unelevated identity. The second matters — collapsing either into a blanket job-token refusal would stop ordinary users creating tokens, and no other case would catch it. The doc comment records which of the seven each route covers. * docs: correct which job-token gates have no observable route The previous commit put `forbid_elevated_job_token` among the guards reachable only through workspace-less routes, and its message named three call sites. It has six, and two are workspaced: `mint_app_embed_token` and `mint_raw_app_sdk_token`. Its superadmin branch is therefore already exercised over HTTP — the 401 the embed-token case asserts is this gate. So three of the seven lack an observable route, not four. Its direct assertions stay: the embed-token case only ever reaches it with an elevated identity, and the unelevated-negative case is what would catch the gate being collapsed into a blanket job-token refusal. * test: pin is_instance_admin, and stop enumerating gates in prose `is_instance_admin` is `authed.is_admin && authed.job_id.is_none()`, so a census built by searching for `job_id.is_some()` could not see it. Both its call sites are workspace-less, and it returns a bool that selects obfuscation rather than refusing — a job token reading `true` leaks `env_vars_static` instead of being turned away. Pin both directions. The doc comment tried to account for every job-token guard and which route exercised it. It was wrong three times running: the count, the call sites of `forbid_elevated_job_token`, and the claim that the CUSTOM_INSTANCE_DB case covers `is_super_admin_authed` when that path tests `job_id` inline. A table that has to be rederived from six crates to stay true does not belong in a comment, so it now states only why these calls are direct. * docs: name the right is_instance_admin caller The comment credited "the concurrency-group listing" with obfuscating rows. The second caller is `prune_concurrency_group`, which returns PermissionDenied; the obfuscating one is `list_worker_groups`. Keep the claim to that single caller, which is what makes this guard fail by leaking `env_vars_static` rather than by admitting a request. --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Ruben Fiszel
|
3c8e4b43fd |
fix: resolve a script path to its new version as soon as the lock lands (#10794)
* fix: resolve a script path to its new version as soon as the lock lands Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011H5ygpzQHkPeYsjiP9GzBy * fix: tell MCP script deploy callers to stop polling on a lock error Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011H5ygpzQHkPeYsjiP9GzBy --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
8b1485a20a |
test: let the benchmark hub serve integration metadata
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |
||
|
Guilhem
|
449b1a6933 |
fix: ground the chat's AI agent provider in the workspace's models (#10774)
* fix: ground the chat's AI agent provider in the workspace's models Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: never reject an AI agent model the catalog could not confirm Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * test: benchmark AI agent provider grounding in ai_evals Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: only reject an AI agent model an exhaustive listing rules out Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep instance-level AI settings out of the workspace provider catalog Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep untrusted model ids out of the chat's context Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * refactor: carry completeness on the model listing itself Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * docs: correct two comments left behind by the catalog rework Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: bound the model listing and verify the default against it Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: honour a workspace default a filtered listing names Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: recognise a workspace default past the prompt's model cap Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: keep an aliasing provider's unlisted model ids permissive Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
8e508ea01a |
feat: support application default credentials for gcp pub/sub triggers (#10778)
* feat: support application default credentials for gcp pub/sub triggers Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: address review findings on gcp application default credentials Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: address review nits on gcp application default credentials Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: key the gcp credential-mode permission off the loaded mode Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: gate enabling an ADC gcp trigger on workspace admin Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: lock the gcp trigger row while authorizing a mode change Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: skip admin-only gcp listing when the caller cannot use those credentials Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore: update ee-repo-ref to 54bf630681000c8ed87a7067e357118e015123b1 This commit updates the EE repository reference after PR #738 was merged in windmill-ee-private. Previous ee-repo-ref: 91d0e228a0ad226625278b400c64f96a61404a10 New ee-repo-ref: 54bf630681000c8ed87a7067e357118e015123b1 Automated by sync-ee-ref workflow. --------- Co-authored-by: Claude Opus 5 <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> |
||
|
Guilhem
|
28b2ca6367 |
feat: inline login errors and a narrower single-column login card (#10777)
* feat: inline login errors and a narrower single-column login card Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address login review findings (overflow, error leak, a11y, dev gate) Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: scope login form ids per instance Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: drop the duplicate dark mode toggle and tighten the login heading gap Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: replay the login shake on every retry, not just the first Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address standards and spec review findings on the login page Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: put the login error under the field it is about Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: attribute a login failure to the credentials it was sent with Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: hide the third-party toggle once the password form is open Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: brand the logged-out pages from one top header instead of a centered logo Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: remember the login method that last worked on this browser Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat: lead the login card with the last used method and anchor its layout Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: address review round findings on the login card Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix: key third-party buttons by method kind and drop a history comment Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> Co-authored-by: Ruben Fiszel <ruben@windmill.dev> |
||
|
hugocasa
|
92a454b7a8 |
fix: split the MCP script tools into createScript and updateScript (#10783)
* fix: let the MCP createScript tool deploy without a parent hash The tool advertised creating a new script with `parent_hash` left unset, but `parent_hash` was one of its declared arguments — and a client that requires every declared argument to be filled has no way to leave it unset. The values such a caller invents (`""`, `"0"`, a zero hash) are all rejected by `/scripts/create`, so no script was ever created. `parent_hash` is now gone from the tool, and the MCP layer sends `auto_parent` in its place: the server resolves the lineage from the path, creating the script when the path is free and deploying a new version of it when it is not. That is what the tool already claimed to do, and it no longer asks the caller to track a hash to do it. `x-mcp-tool-fixed-fields` is the general mechanism behind this — body fields the MCP layer fills in itself, absent from the tool schema. A null argument is also dropped from the assembled body now, for the same reason the placeholder hashes were a problem: it is how a caller with no value to give says so, and the API rejects it rather than falling back to the field's default. Fixes GIT-973 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: hash a script version once auto_parent has resolved its parent `create_script` hashed the incoming script before the `auto_parent` block filled in `parent_hash`, and the version hash covers that field. A deploy that let the server resolve the parent was therefore hashed as if the path had no history, so redeploying content the path had held before collided with that archived version and returned "A script with same hash ... already exists!" instead of becoming a new version of the lineage. Reverting a script to an earlier state was impossible for any caller relying on auto_parent alone, which is now every MCP caller. The hash and the duplicate-hash check move below the resolution, so an auto_parent deploy hashes the lineage it will actually be attached to. Callers passing an explicit `parent_hash` are unaffected: the resolution block leaves their `ns` untouched, so they hash exactly as before. The CLI masked this by sending `parent_hash` and `auto_parent` together, using auto_parent only as a stale-hash fallback. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * docs: state the constraint that pins the script hash site Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: reject a fixed-fields spec the MCP layer would not honour `validate_fixed_fields` ran only for an operation that declares a request body, and passed any body whose properties it could not see. Two shapes reached the generated tool with fixed fields that are dropped at call time: an operation with no `requestBody`, where the body builder returns before reading them, and a pass-through body, which carries the runnable's own arguments and never receives a key of ours. Both are now generation-time errors, so the only specs that get the extension are the ones where it means something. Also name the folder-derived `on_behalf_of` alongside `parent_hash` at the hash site: both are written to `ns` before it, and a reader who knows about only one could reintroduce the early hash. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix: keep fixed fields internal and catch a misspelled one `EndpointTool` is what `list_tools` publishes as the tool catalogue, so deriving `body_fixed_fields` into it put a field in the caller's view that is by definition not the caller's to set, and that the OpenAPI schema does not declare. It is no longer serialized. The generator also only checked a fixed key against the exposed subset of the body properties, which cannot tell a field deliberately left out of `x-mcp-tool-include-fields` from a misspelling of one. A key the API does not declare is now a generation-time error rather than one serde discards in silence, and the extension must be a non-empty mapping — an empty list previously slipped through the type check on its way to being ignored. Narrow the hash-site comment to the ordering it actually constrains. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat: split the MCP script tools into createScript and updateScript Scripts were the only entity in the MCP surface without the create/update pair every other one has, because the REST API has no update route for them: a script is immutably versioned, so `POST /scripts/create` is also its update, and one tool had to infer which the caller meant from the state of the path. That inference is what GIT-973 is. `parent_hash` told the two apart, and an MCP client that requires every declared argument to be filled has no way to leave it unset, so no script could be created: `""` is a 422, `"0"` is a 422, and `"0000000000000000"` is a 400. Naming the intent removes the field instead of the guard. `createScript` means the path should be free and keeps refusing an occupied one; `updateScript` names the version it supersedes in its URL, so the body carries no hash either. Picking the wrong one now fails loudly rather than succeeding on the wrong script. - New `POST /w/{workspace}/scripts/update/{path}`, deploying a new version of the script the URL names. Its body `path` is the destination, defaulting to the URL's, so setting a different one moves the script and keeps its history — which no MCP client could ask for while `createScript` was the only tool. - New `x-mcp-tool-optional-fields`, dropping a body field from the tool's `required` where the handler defaults it. `updateScript` uses it for that destination path: required, an agent has to restate the path on every edit, and a value that drifts from the URL's silently moves the script. - `assemble_request_body` drops null-valued arguments, matching what the pass-through branch already did. A client that must fill in every argument says "no value" with `null`, and the API rejects that for a bare `String` field rather than falling back to its default. Fixes GIT-973 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * fix: confine updateScript to the token's script paths `endpoint_path_policy` is what applies an `mcp:scripts:<pattern>` token's path patterns to an endpoint tool, and a tool it does not name is not confined at all. `updateScript` was not named, so a path-scoped token could deploy over, and move, any script in the workspace: the proxy mints a bare `scripts:write` for a caller whose only scopes are `mcp:`-prefixed, and nothing downstream held a pattern. The destination path has to bind only when supplied — omitting it is how a caller updates in place — so `PathArgs` grows `optional_fields`, checked when present and never required. Empty reads as absent, matching the handler, which now takes an empty body `path` for "leave it where it is" rather than moving the script to the empty path: a caller obliged to fill in every field sends `""` as readily as null. That shape also fixes `updateFlow`, whose entry named `path__path` for the URL argument. The generator gives the URL path the plain name, so the lookup never matched and every confined call failed closed on a missing argument. Both sides now have a drift guard: a script/flow tool the URL addresses by path must have a policy. The backend one lives in windmill-api, where the generated catalogue is, since the policy is in windmill-mcp and neither crate sees both. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * fix: address the review round on the script tool split Four findings, three of them one bug: a destination path the caller left empty. `update_script` read it as "leave it where it is", the confinement check skipped it on the strength of that, and `update_flow` did neither — it takes the empty string literally and moves the flow there, so the skipped check was the only thing standing in front of that move. A database constraint refuses the empty path, so nothing was reachable through it, but the confinement was relying on a property of one handler that its sibling did not have. The MCP layer now strips an empty optional destination from the arguments, so no handler receives one and there is nothing left for the check to skip. Neither tool depends on the other's reading of it any more. `update_script` also resolved the head before opening the deploying transaction. A version landing in between is caught — it leaves a child behind, and the linear-lineage check refuses that — but an archive leaves none, and the hash of an archived version still exists, so the deploy would have chained onto it and revived the script the archive had just retired. The resolution moves into the transaction. The scope check on the URL path moves ahead of that resolution, so a path outside the token's scope answers the same whether or not a script is there, rather than telling the two apart through 404 against 403. `x-mcp-tool-optional-fields` goes: the generator already strips a body field that collides with a same-named path parameter from `required`, so the extension regenerated byte-for-byte identical output. The test that pinned the destination as optional stays — it pins the behavior, which is now the collision handling's. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * fix: lock the head an updateScript supersedes Moving the resolution into the deploying transaction narrowed the archive race without closing it. The plain SELECT took no row lock, so an archive could still land between it and the parent-existence check below, which finds the parent by hash and never looks at `archived` — the deploy then chained onto the archived version and inserted a live child, reviving the script the archive had retired. `FOR UPDATE` on the resolution is what makes the row the head rather than a head it once was: the archive either waits for the deploy, or wins and leaves the row failing the `archived` qualifier on re-check, so no version resolves at all. The regression test stages that interleaving rather than approximating it. It holds the head row from a second connection so the deploy parks on it, waits for a backend to actually be blocked before archiving — without that wait the request loses to a local UPDATE and never reaches its resolution, which is the sequential case the neighbouring test already covers — then asserts the update is refused. It returns 201 and revives the script with the lock removed. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * fix: have the MCP layer name the path an update keeps The tool lets a caller omit the destination, and the endpoint was absorbing that by accepting a body without a `path` and defaulting it from the URL. The OpenAPI schema says `path` is required, so the two disagreed and a generated REST client could not follow the contract the description promised. The MCP layer fills the destination in instead, from the path the item is already at, since that is what omitting it means. The endpoint then always receives a body naming its own path and matches its schema, `update_script` takes a `NewScript` rather than picking a JSON object apart to inject a default, and the empty string stops being a value any handler has to interpret — `update_flow` reads one as the empty path, which is why it was stripped a commit ago. The alternative, an `EditScript` schema differing from `NewScript` only in whether `path` is required, was measured and rejected: openapi-ts drops the `required` of an `allOf` branch, so `NewScript` came out with every field optional and broke 15 frontend types. Loosening a schema every API consumer shares, to make one field optional on one route, is the worse trade. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * fix: tell a superseded update apart from a missing script Locking the head made the loser of two concurrent deploys answer 404 "Script not found" for a path the caller can see holds a script: its lock re-check finds the row archived and filtered, and nothing looked further. It now looks — a live version at the path means this deploy lost to one that superseded the version it set out to supersede, which is a conflict to retry, not a script to go find. The regression test stages that interleaving the way the archive one does, with the winner leaving a live head behind rather than an archived path. It answers 404 with the branch removed. The rationale for the lock also sat in two places; it stays at the query, which is where dropping it would do the damage. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * docs: drop the path default update_script no longer applies The handler stopped defaulting the body's path when the MCP layer took the job over; its doc comment still described the old contract. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * docs: sync the deref YAML with the update route's path contract The dereferenced bundle rewraps prose at its own width, so the edit that updated the canonical spec and the JSON bundle matched nothing here and left the served YAML still offering a default the endpoint no longer applies. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * docs: stop the script tools describing a parent_hash they cannot take `description` is read by two audiences: it documents the route, and it opens the MCP tool's text. Written for the first, it told an agent that createScript "does it too when given that version's `parent_hash`" — a field neither tool exposes, and inviting exactly the call this branch exists to make impossible. updateScript's told the agent to repeat the URL's path while its own instructions say to omit it; both work, since the MCP layer fills it in, but only one of them can be the advice. Both now describe what the operation does and leave the mechanics to the text that belongs to each caller: the request body's own description for REST, the tool instructions for an agent. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * docs: give the create route's two audiences their own description Removing the `parent_hash` sentence took a true fact out of the REST documentation: the create route does still deploy a new version, and still rename, when the body names the version it supersedes. Nothing replaced the explanation, and the field carried no description of its own. `description` cannot serve both readers — it documents an endpoint whose schema has `parent_hash`, and it opens a tool whose filtered schema deliberately does not. `x-mcp-tool-description` stands in for it on the tool, the way `x-mcp-tool-name` already does for the name, so the route keeps its full contract and the agent is not told to send a field it has no way to send. What `parent_hash` does now sits on the field, where a REST caller looks for it and where `x-mcp-tool-include-fields` drops it before an agent sees it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * docs: tell an agent a new version is not runnable the instant it deploys A deploy returns before its lockfile exists, so a script run straight after one can still execute the previous version. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * docs: say why a new version is not runnable the instant it deploys Its lock is generated asynchronously, so a script run straight after a deploy can still execute the previous version. On both script tools: a freshly created script is no more immediately runnable than a freshly updated one. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s * docs: bound the wait after a deploy instead of naming a signal for it `getScriptByPath` reports the new hash the instant the version exists, while its lock is still null, so the previous version is what a run by path executes. There is no signal that fixes this: the deploy evicts DEPLOYED_SCRIPT_HASH_CACHE, but anything resolving the path before the lock lands re-populates it with the old hash, and the lock landing evicts nothing. Waiting for a non-null lock is necessary and not sufficient, so pointing at one would have been a second wrong answer. Measured: a run right after the lock lands still gets the previous version, and the same run 65s later gets the new one, which is the cache's 60s TTL. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CEWFHpmTBauDBi93MnsT7s --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> Co-authored-by: Ruben Fiszel <ruben@windmill.dev> |
||
|
Guilhem
|
1a7682891d |
collapse conditionally hidden fields in schema forms (#10791)
A field whose `showExpr` evaluates false skipped its `ArgInput` but still rendered the padded row that wraps it, so the enclosing ResizeTransitionWrapper measured 8px (16px with `largeGap`) of leftover padding per hidden field. Simulating a `oneOf` with a selector and one `showExpr` branch per variant stacked one such gap per unselected branch. Move the `!hidden[argName]` check onto the row itself so nothing is rendered for a hidden field and the wrapper collapses to 0px. Claude-Session: https://claude.ai/code/session_01Lx7KEPQC4SXjVjYkkks7Za Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Guilhem
|
c7e3537da8 |
give every brand icon the lucide safe area and centre its artwork (#10790)
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
hugocasa
|
2ca2fecce9 |
fix: treat hub curated as three-state and speak only for a stated true
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> |