13 Commits

Include renames

Author	SHA1	Message	Date
Waleed	47f143016e	fix(docs): restore api-reference URL continuity and fix translated SDK bodies (#6617) ... Two docs-only defects introduced by #5273 (`263e3ca67e`), which re-founded the public API reference on the v2 surface. 1. Ten translated SDK snippets produce a deterministic 400. The streaming example in the five translated `api-reference/typescript.mdx` and `python.mdx` pages was repointed from `/api/workflows/{id}/execute` to `/api/v2/workflows/{id}/execute` and nothing else was changed — fr/ja/zh typescript.mdx are literally one-line diffs. `message` stayed at the body root. That was correct against v1, whose route treats the whole non-control body as workflow input, but `v2ExecuteWorkflowBodySchema` ends in `.strict()` and the route parses before executing, so every copied snippet returns `400 Unrecognized key: "message"`. The same commit fixed the English bodies to `input: { … }`, so this is an oversight, not a decision. The ten fences now match `en/api-reference/typescript.mdx:959` and `python.mdx:681`. Not relaxing `.strict()`: it is deliberate house style across the v2 contract and is what makes a typo'd option fail loudly instead of silently. 2. Thirty-two published operation pages 404 with no redirect. Replacing the single v1 `openapi.json` with seven v2-only specs changes page identity, because fumadocs derives every generated page as `slugify(tag)/operationId` from the specs at build time. Re-deriving both sets gives 52 old slugs and 128 new ones: 32 disappear and 20 keep their URL while silently retargeting v1 -> v2 (`knowledge-bases/updateKnowledgeBase` also flips PUT -> PATCH). All 52 are in the live sitemap — parsing `<loc>` from docs.sim.ai/sitemap.xml gives 458 URLs of which 56 are `/api-reference/`: the four static pages plus all 52 generated ones by name, including every one of the 32 that die. They are 200 today under an allow-all robots.txt. The spec swap itself is deliberate and CI-enforced (`check-openapi-specs.ts` requires every published operation under `/api/v2/`), so restoring the v1 operations is not an option. The missing piece is the redirect map, in a file that already carried 56 such rules from earlier doc moves. `permanent: true` (308) is used only for a true 1:1 successor — same operation, renamed. A 308 is cached indefinitely and effectively unrecallable, so anything that collapses two pages onto one, changes the identifier model, or lands on a merely adjacent operation is `permanent: false` (307). That splits 21/11. Four destinations differ from the mapping proposed in review, each on evidence from the specs rather than from the operation names: - `workflows/getJobStatus` is not destination-less. The v2 queued-execution receipt (`QueuedWorkflowRun`) returns `statusUrl` `/api/v2/workflows/{id}/runs/{runId}`, so `workflow-runs/getWorkflowRunV2` is the successor poll target — far better than a generic landing page. - The three HITL read operations go to `getWorkflowRunV2`, not to the resume page: `WorkflowRunStatus` carries a `paused` object with `contextId`, `pausedAt`, and `pauseKind`. Pointing a GET doc at a POST doc would be wrong. - `human-in-the-loop/listPausedExecutions` goes to `listWorkflowRunsV2`, whose `status` filter includes `paused`. - `tables/batchUpdateRows` is 307, not 308. v2 `updateTableRows` is "Update Rows by Filter" — the successor of v1 `updateRows` (PUT, predicate-based), which keeps its 308. v2 has no by-id batch update at all, so batchUpdateRows lands on a genuinely different operation. 3. A guard, so the map cannot rot silently. `scripts/openapi/docs-redirects.test.ts` recomputes the generated slug set the way fumadocs does and asserts no `/api-reference/` source shadows a live page and every destination resolves. Nothing else in the repo reads docs URLs, so a future spec regeneration would otherwise break the map with no signal. It needs no wiring: `check-openapi.ts` already runs this vitest config. The redirect array moves to `apps/docs/lib/redirects.ts` because the guard cannot import `next.config.ts` — `createMDX()` runs the fumadocs-mdx generator at import time, which made vitest emit an unhandled build error and warn about false positives. The 56 pre-existing rules are byte-identical to before, verified programmatically; `next.config.ts` keeps the same public shape and Next's own `checkCustomRoutes` accepts all 88 rules. Open question for the owner, larger than the redirects: all 82 `/api/v1` route files survive on staging, so a live public API now ships with zero reference docs, while the documented `/api/v2` surface returns 404 for any caller outside the off-by-default `v2-api` flag cohort. Is that the intended end state or transitional?	2026-08-12 01:49:59 -07:00
6 people	263e3ca67e	improvement(external-endpoints): v2 versions with clean signatures + updated docs based on openapi spec (#5273) ... * v0.6.29: login improvements, posthog telemetry (#4026) * feat(posthog): Add tracking on mothership abort (#4023) Co-authored-by: Theodore Li <theo@sim.ai> * fix(login): fix captcha headers for manual login (#4025) * fix(signup): fix turnstile key loading * fix(login): fix captcha header passing * Catch user already exists, remove login form captcha * improvement(external-endpoints): v2 versions with clean signatures + updated docs * feat(usage): accept X-API-Key on usage-logs list + export /api/users/me/usage-logs and /export now use checkHybridAuth — the same auth /api/users/me/usage-limits already accepts — so external monitors can read summary.bySourceCredits (the source breakdown of usage-limits' aggregate currentPeriodCost) instead of estimating Copilot spend by subtraction. Workspace-scoped keys are pinned to their own workspace's slice of the ledger: the filter defaults to the key's workspace and an explicit mismatch 403s. Both endpoints documented in openapi-core.json. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(billing): dedicated v2 usage endpoints; keep internal usage routes session-only Replaces the earlier X-API-Key enablement on /api/users/me/usage-logs with a dedicated public surface, so the internal Billing-settings endpoints can evolve with the UI while external monitors get a stable versioned contract: - GET /api/v2/billing/usage — current-billing-period summary with bySourceCredits (the source breakdown external monitors need to watch e.g. Copilot consumption without estimating by subtraction), plus limitCredits and plan - GET /api/v2/billing/usage/logs — cursor-paged credit ledger in the v2 envelope - workspace-scoped keys are pinned to their own workspace's slice; personal keys read the account ledger The public wire is credits-only: usage-logs rows now carry a hasCost boolean instead of dollarCost (the Billing UI only needed the >0 signal), and the rateLimit block is removed from the usage-limits response and docs (deploy-modal tab relabeled accordingly). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(docs): validate OpenAPI specs against the Zod contracts in CI The specs in apps/docs are hand-authored because they carry what Zod never defines — error envelopes, status codes, prose, examples — so they can't be generated; check:openapi validates them instead: - spec integrity: $refs resolve, operationIds unique, 2xx documented, no orphaned component schemas - v2 conventions: every /api/v2 operation documents 401 + 429 and every 4xx/5xx resolves to the canonical { error: { code, message } } envelope - contract cross-check: contracts are auto-discovered from lib/api/contracts/v2 (each carries its method + path); doc<->contract coverage both ways, query/body/response field diffs via z.toJSONSchema - examples: documented request/response examples must parse with the matching contract's actual Zod schemas First run caught real drift, fixed here: 16 stale orphaned schemas in the core spec, the v2 billing ops referencing v1-shaped error components, deploy/rollback examples missing the required nullable lifecycle keys, CreateTableBody missing folderId, a legacy-grammar delete-rows example, and four knowledge document ops missing their required workspaceId query param. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * fix(docs): recursive field diff in check:openapi + the deep drift it found A mutation test showed the doc<->contract field diff only compared top-level properties, so a typo inside the { data } envelope passed. The diff now descends through matching object properties and array items (both sides must expose a property set — passthrough contracts and prose-only docs end the descent instead of false-positive), with the Zod JSON-schema root doubling as the $defs context. Deep drift it immediately caught, fixed here: select-column config (options/multiple) missing from every tables column schema, AddColumnBody hand-rolling a third column shape (now composed from ColumnInput, with position/workflowGroupId as the per-op extensions the contracts actually admit), chunking strategyOptions undocumented, and the deployment lifecycle fields (activeDeployment/latestDeploymentAttempt) missing from DeploymentState. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * fix(security): close the triggerType rate-limit bypass on workflow execute Caller-supplied triggerType flowed unchecked into preprocessExecution, whose checkRateLimit default turns OFF for 'manual'/'chat' — so any API-key caller, and any anonymous public-API caller billed to the workspace owner, could execute unthrottled by sending {"triggerType":"manual"} (async runs also skipped the worker-side check via admissionCompleted). External callers may now only send the redundant 'api' value; internal JWT callers ('workflow'/'mcp') are unaffected. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * refactor(execution): extract enqueue/status/cancel into shared libs Prepares the v2 execution surface: handleAsyncExecution's queue logic moves to lib/workflows/executor/enqueue-execution.ts (slot/claim semantics encoded in a discriminated outcome, not HTTP statuses), the execution-status read to execution-status.ts, and the order-sensitive cancel machinery to lib/execution/cancel-workflow-execution.ts. The v1 routes re-render identically — their suites pass unmodified. Also: preprocessExecution gains rateLimitCounter ('sync'|'async') and its 429 now carries code RATE_LIMIT_EXCEEDED + retryAfterMs (previously indistinguishable from the concurrency 429 and Retry-After was discarded); and the duplicate cancel contract in contracts/logs.ts is unified on the full 5-value reason enum — its narrower copy made requestJson throw a client ZodError when cancelling a paused HITL run. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(execution): callable execution service + structured error classifier executeWorkflowService composes the same libs the v1 route holds inline (call-chain guard, execution-id claim, LoggingSession, preprocessing, deployed-state load + file-field processing, timeout-bound executeWorkflowCore, output hydration/compaction) for the deployed-state caller class — the seam the v2 execute route and in-process internal callers share, making the HTTP endpoint syntactic sugar. classifyExecutionError stops discarding the block context that buildBlockExecutionError already attaches at throw sites: failed runs now yield {message, code, blockId, blockName, blockType} with a stable append-only code enum (TIMEOUT/CANCELLED/USAGE_LIMIT_EXCEEDED/ INVALID_INPUT/BLOCK_EXECUTION_FAILED/CHILD_WORKFLOW_FAILED/ OUTPUT_TOO_LARGE/EXECUTION_FAILED), so callers route on error class instead of substring-matching messages — the single place raw errors are interpreted. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(api): POST /api/v2/workflows/[id]/execute Thin route over executeWorkflowService: X-API-Key or anonymous public-API auth (sync/stream only for anonymous), strict body with body-flag async (no mode headers on v2), SSE passthrough for stream, and the execution resource response — executionId always present, in-band run failures are status:'failed' with the structured {message, code, blockId, blockName, blockType} error, sync timeout is status:'failed' + TIMEOUT instead of v1's 408, and a Response block's payload stays inside output (authors never control response status/headers on this origin). Async debits the async bucket and the 202 statusUrl points at the v2 executions resource. Adds CLIENT_CLOSED_REQUEST/SERVICE_UNAVAILABLE to the v2 error codes. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(api): v2 executions status + cancel with queued backfill GET /api/v2/workflows/[id]/executions/[executionId] is the single status URL for sync and async runs: before the async worker writes the durable log row, status is backfilled from the job queue (deterministic job id) as 'queued'/'running' — closing v1's 202-to-pickup 404 window — and failed runs carry the structured error object. POST .../cancel renders the shared cancellation lib in the v2 envelope with the tightened 5-value reason enum. Both authenticate via the shared resolveV2WorkflowAccess (X-API-Key, authz masked as 404, allowPersonalApiKeys honored). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(execution): workflow tool + MCP bridge run in-process workflow_executor (workflow-as-agent-tool) short-circuits in executeTool through WorkflowBlockHandler — the same invocation boundary canvas child workflows use — mirroring the deployed_block_executor precedent. The MCP serve bridge calls executeWorkflowService directly instead of fetching its own execute endpoint; deployment-version pinning, MCP response-size rejection, and the actor override become typed options instead of header sniffing. Both callers drop the double admission slot and duplicate top-level log row the HTTP hop cost, and failed child runs now surface the structured error + child executionId so parents and MCP clients can route on error class and hand providers a reproducible handle. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(infra): CORS + CSP coverage for the v2 execute path /api/v2/workflows/:id/execute gets the same wildcard-origin, credential-free CORS policy as v1 (the default credentialed policy would block browser API-key calls and open a cookie CSRF surface) with X-Sim-Stream-Protocol allowed and no X-Execution-Mode (async is body-selected on v2), plus the COEP/COOP/CSP header block. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(ui): deploy modal + copilot advertise the v2 execute surface All 20 API-tab snippets move to POST /api/v2/workflows/{id}/execute with the nested {"input": ...} body, async as the "async": true body flag (X-Execution-Mode gone), status polling against the v2 executions resource, the third tab renamed Usage and pointed at /api/v2/billing/usage, and {data} envelope unwraps in the printed responses. Fixes the latent baseUrl derivation (endpoint.split('/api/workflows/')) that would have silently built garbage URLs under a v2 endpoint, and deletes dead code (exampleCommand across 3 sites, getAsyncExampleTitle). Copilot deploy/manage/serializer endpoint builders and the api_trigger bestPractices example follow (the latter also drops its hardcoded staging host). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * docs(api): document the v2 execution surface Adds execute, execution status, and cancel to openapi-v2-workflows.json with the structured ExecutionError schema (append-only code enum + block attribution) and the ExecutionResource contract, documenting the rules that differ from v1: modes are body-selected, a failed run is HTTP 200 with status 'failed', an executionId always means data (never the error envelope), queued status is visible immediately, and Response-block payloads stay inside output. Registers the three pages in the generated workflows meta.json and bumps the route-count baseline. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * feat(api): gate the whole /api/v2 surface behind one flag; UI stays on v1 Every v2 route now runs exactly one check immediately after auth — v2ApiGateError — and answers 404 when the `v2-api` flag is off, so the surface is invisible until it is deliberately rolled out. The gate is keyed on userId only: a workspace/org-keyed check would have to read membership for a caller-supplied id before authorization runs, and its 404-vs-403 split would leak cohort membership (the trap the per-domain table gate worked around by running late). The two executions routes inherit it from the shared access resolver; the tables-specific gate is removed so no route checks twice. `tables-v2-api` stays, now gating only the internal predicate-grammar route /api/table/[tableId]/query — note v2 tables routes move to the unified flag, so enabling them is a `v2-api` decision now. Reverts the deploy modal, copilot handlers, and api_trigger example to the v1 execute endpoint: v1 works unchanged, and the UI must not advertise a surface most users would get a 404 from. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CiHhAk2R1NryaS3R8n2yFz * fix(executor): restore child-cost aggregation dropped by the staging merge Staging's custom-block rewrite deleted `aggregateChildCost` from workflow-handler.ts, and git merged that file cleanly — but this branch's workflow-tool-runner.ts, added for the v2 execute migration, still imports it. A silent semantic conflict: no marker, broken build. Taking staging's rewrite is correct, so the helper is defined locally in its one remaining consumer rather than resurrected in the file staging just rewrote. Same four lines over the still-exported `calculateCostSummary`, so a failed child workflow keeps billing the hosted-key spend it consumed instead of reporting $0. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * refactor(tables): make lib/table/orchestration the single implementation (#6134) * refactor(orchestration): move the shared error contract out of lib/workflows OrchestrationErrorCode and statusForOrchestrationError are the contract every lib/[resource]/orchestration module returns against, but they lived inside the workflows module, so resource-neutral code (lib/folders) already had to import from a workflow path. Moved to lib/core/orchestration/types. Adds a 'locked' class mapping to 423. Both tables and workflows have a lock that forbids a mutation, and each caller was translating that to a status itself. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * refactor(tables): make lib/table/orchestration the single implementation Column update was implemented four times — the UI route, v1, v2, and the copilot table tool — each calling the same column services but owning its own guards, error mapping, and audit. The copies had drifted, and the drift was the bug: v2 was missing both guards, only the copilot copy minted stable option ids, and only v1/v2 audited. performUpdateTableColumn, performDeleteTable, and performDeleteTableRow now own that logic; all ten call sites reduce to auth, parse, call, render. The guards are asserted once in lib/table/orchestration rather than four times against four routes. Behavior this consolidates, previously true on only some paths: - The typeChanging guard. updateColumnType early-returns on an unchanged type and drops any options sent with it, so restating the current type alongside new options silently discarded them. v2 had no guard at all and, since its contract shares v1's body schema, accepted options and ignored them. - The select-unique guard. Each write is its own locked transaction, so a rename or type change paired with a constraint write that is going to fail commits first and then throws, half-applying the schema change. - Stable select-option ids. Cells reference the option id, so an edit that re-sends an option by name has to reuse it or every cell holding it is orphaned. Only the copilot path did this; normalizeSelectOptionsInput moves to lib/table/select-options and now covers every caller. It preserves a supplied id, so it is a no-op for the fully-formed options the HTTP contracts accept. - required forwarded into the type and options writes, so a conversion validates against the constraint the same request is setting. - An audit on every successful update. The UI route and the copilot tool emitted none. - Single-row delete through the row service. v2 did a raw db.delete, skipping assertRowDelete and deleteOrderedRow, so a delete-locked table returned 200 and the row-count bookkeeping never ran. - The delete actor handed to deleteTable, which audits only when a row was actually archived. v1 and v2 omitted it and audited themselves outside that check, emitting TABLE_DELETED for a no-op delete of an archived table. Failure classes come back as OrchestrationErrorCode; v2 renders them through a new v2ErrorForOrchestration, mirroring statusForOrchestrationError on the v1 and UI surfaces, so a given failure maps to the same status everywhere. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test(tables): bind the column-update tests to the orchestration function The base's route tests assert which column service each payload reaches — the behavior that now lives in performUpdateTableColumn. They mocked the `@/lib/table` barrel; the orchestration module imports the service directly, so they mock that too and keep asserting the same thing through the extracted implementation. The orchestration tests move onto the base's semantics: writes address the stable column id, a rename rides inside the write it accompanies rather than running first, and the currency guards replace the non-select options guard the service now owns. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore(copilot): drop the column-type import the delegation made dead Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * refactor(tables): move the audit log out of the table service `lib/table/service.ts` wrote its own audit rows, so whether an operation was audited depended on which function a caller reached for rather than on a user having performed it. That is what let v1 and v2 audit a no-op delete, and what made `deleteTable`'s optional `actingUserId` double as an audit opt-out flag. Worse, most sites fell back to `actingUserId ?? createdBy`, so an unattributed call was logged against the table's *creator*. The copilot `mv` path passed no actor at all: renaming someone else's table recorded them as the renamer. Audit now lives in the orchestration functions — performDeleteTable, performRenameTable, performMoveTableToFolder, performUpdateTableLocks — and the services just write. Internal callers (folder cascade, import rollback) keep calling the service and are silent by construction rather than by remembering to omit an argument. Two services now return what the audit needs: `deleteTable` reports whether it actually archived a row, so a repeat delete logs nothing; `updateTableLocks` returns the before/after locks, since only the locked write can observe the transition its description names. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(tables): restore audit provenance and conflict status in orchestration Moving the audits into the orchestration functions dropped three things the routes had been carrying, and added one the orchestration now owns twice. - The v1 and v2 column-update routes passed `request` to `recordAudit`, so their audit rows recorded the caller's IP and user-agent. The orchestration function had no way to receive it. Every table orchestration function now takes an optional `OrchestrationRequestContext` and every HTTP route forwards it; the copilot and VFS callers, which have no request, omit it. - `classifyTableMutation` matched `TableConflictError` on "already exists" appearing in the message and reported it as `validation`, turning the UI route's 409 on a duplicate table rename into a 400. It now matches the type, the way `performRestoreTable` already did. - `captureServerEvent` ran on every delete while the audit was gated on a row actually being archived, so a repeat delete of an archived table still reported `table_deleted`. Both now hang off the same evidence. - The copilot delete path kept its own `captureServerEvent` from when the service did not emit one, double-counting every copilot table delete. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01YGzbVDZpe2dEALbu2BUU8a * fix(tables): say which type a no-op column update restated A copilot `update_column` payload whose only content was the column's current type used to return success with the live schema, while the v1, v2, and UI routes rejected the same payload with "No updates specified". Delegating to `performUpdateTableColumn` unified them onto the routes' rejection — correct, but the message tells the caller its request was empty when it named a type. The orchestration function now reports the same thing `updateColumnType` reports when it loses this race concurrently: the column is already that type, re-issue without the type change. An empty payload still reads "No updates specified". Drops the copilot's `outcome.table ?? tableForUpdate` fallback with it — the comment described the no-op that can no longer reach that line, and a success always carries a table. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01YGzbVDZpe2dEALbu2BUU8a * refactor(tables): classify failures by type instead of by message text The table module decided HTTP statuses by searching error messages for phrases. `VALIDATION_MESSAGE_FRAGMENTS` and `ROW_WRITE_ERROR_PATTERNS` held 32 substrings between them, and fifteen more lists were inlined in routes — 83 matchers over 17 files, each its own copy of the guesswork and already drifted apart. It made message wording load-bearing: `TableRowLimitError`'s own doc comment noted that its text had to contain "row limit" for a route to answer 400, and adding "already exists" to a rename message silently demoted a 409 to a 400 (the bug fixed one commit ago, by adding another special case). Services now throw `OrchestrationError`, which carries the transport-neutral `OrchestrationErrorCode` the layers above already speak. Classification is one `instanceof` in `orchestrationErrorResponse` (UI + v1) and `v2CaughtOrchestrationError` (v2). Every pattern list is gone. Wording is free to change; an unclassified error still becomes a generic 500, which is what an unexpected fault should be. `asOrchestrationError` walks the `cause` chain rather than testing the caught value directly: drizzle wraps a throw raised inside a transaction callback in a `DrizzleQueryError` whose own message is the failed SQL, so a bare `instanceof` would drop every failure raised inside `withLockedTable`. That is the same reason `rootErrorMessage` had to dig for a root cause before. Three throws stay bare `Error` deliberately — `Table ID mismatch`, `Workspace ID mismatch`, and `Failed to build upsert conflict predicate` are internal invariants no consumer classified, and they keep falling through to a 500. `Insufficient capacity` was in the pattern list with no producer anywhere in the codebase. Status changes, all deliberate: - `'forbidden'` joins the code union so the table-row-limit ceiling keeps its 403; without it this refactor would have flattened it to 400. - import-async's table-limit rejection: 400 -> 403, matching the two other create routes it had drifted from. - Renaming a table to an invalid name: 500 -> 400. `validateTableName` messages don't contain "Invalid", so no matcher ever caught them. - Restoring a table that isn't archived, or into an archived workspace: 500 -> 400. - A duplicate *column* name stays `validation`/400 rather than becoming a 409 like a duplicate table name. Both v1 and the orchestration have always answered 400 for it; changing a published status is not this refactor's job. The twelve tests that changed were asserting the substring mechanism itself, constructing plain `Error`s with magic strings. They now assert the real contract, plus new cases pinning that identical wording carrying no classification stays internal and keeps its message off the wire. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01YGzbVDZpe2dEALbu2BUU8a --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> * feat(api): add v2 endpoints for MCP servers, skills, custom tools, folders, and credentials (#6150) * feat(api): add v2 endpoints for MCP servers, skills, custom tools, folders, and credentials * fix(api): correct credential role, skill permission bar, MCP url identity, and custom-tool conflict mapping * fix(api): align credential mutation gating, provider-outage status, and unique-violation conflicts * fix(api): close unique-violation, revival, orphan-write, and env-rename gaps * fix(api): treat every provider-outage code as unavailable on create and update * fix(credentials): use the shared outage predicate on the session update path * fix(contracts): anchor the predicate double-cast annotation to the cast `check:api-validation:strict` counted 9 unannotated double-casts against a baseline of 8, failing CI. The predicate leaf schema was annotated, but the annotation sat above the declaration while the checker anchors on the line carrying the cast — five lines below, at the close of the object literal. The scanner walks back at most three lines and stops at the first non-comment one, so it hit `value: z.unknown().optional(),` and never saw the reason. Splitting the object schema from the cast puts them adjacent, so the existing reason binds. No behavior change — the cast, the schema, and the reasoning are unchanged. Also lowers the rawJsonReads ratchet 6 -> 5 to match the current count, which had drifted down; leaving it high lets a removed raw read silently come back. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(skills): point the orchestration error contract at its moved module #6150 branched before #6134, so skill-lifecycle.ts imports @/lib/workflows/orchestration/types — the module #6134 moved to @/lib/core/orchestration/types. Git merged a file deletion on one side with a new file referencing it on the other: no textual conflict, broken build. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * refactor(knowledge): make lib/knowledge/orchestration the single implementation (#6154) * refactor(knowledge): make lib/knowledge/orchestration the single implementation Knowledge base create was implemented four times — the internal route, v1, v2, and the copilot tool — and the orchestration around the shared write had drifted. Extract it the same way lib/table/orchestration was: services write, orchestration decides which writes run, guards them, audits them, and returns a transport-neutral failure. Behavior converged, not preserved: - One chunking default (DEFAULT_CHUNKING_CONFIG). The agent defaulted minSize to 1 against the API's 100, so identical input produced differently-chunked knowledge bases depending on who created it. The agent path now chunks at 100. - Every successful mutation is audited inside the orchestration function. The copilot tool called recordAudit zero times, so agent-created knowledge bases, document uploads, updates and deletes left no audit trail at all. - Failures classify by class, not by message text. The knowledge service errors are OrchestrationError subclasses and storage-quota rejections throw a shared StorageLimitExceededError, replacing four separate message greps for "already exists" / "does not have permission" / "storage limit". delete_connector reported the opposite of what happened. It reached the route through an internal HTTP self-call that sent no query string, so the route's keep-documents default always applied while the agent told the user the documents had been removed. The self-call is gone — all four connector operations run in-process — and the orchestration returns the real counts. Also: - OrchestrationErrorCode gains 'payload_too_large' (413 / PAYLOAD_TOO_LARGE). Without it, dropping the storage-limit message match would have regressed the documented 413 on knowledge base create and document upload to a 500. - messageForOrchestrationError renders a route's own wording for an unclassified fault, so a driver's message no longer reaches the client on a 500. - v1 and v2 knowledge base update now forward actorUserId, which the service requires for a workspace move; both omitted it. - The connector DELETE route reads deleteDocuments through parseRequest. Its contract declared z.boolean(), which would have rejected the string a query param actually is. - Drop the 409 from POST /api/v2/knowledge/{id}/documents in the OpenAPI spec. Nothing on the upload path throws a conflict; it was only ever reachable by the message match this change removes. Behavior change worth noting: a v1/v2 PUT carrying only the workspaceId scope field and no actual updates now returns 400 rather than 200 with the unchanged knowledge base. Deliberately deferred: document update remains internal-only. Extracting performUpdateKnowledgeDocument makes exposing it on v1/v2 a contract and a route away, but that is a new public surface rather than part of this consolidation. * fix(knowledge): make connector create atomic and stop flattening failures Review round 1 on #6154. - Resolve the billing payer before the connector is committed, not after. A malformed attribution header rejected post-commit left a live connector behind a 500, and a retry created a duplicate plus duplicate sync work. Manual sync resolves before writing its audit for the same reason. - Let the source-config validator carry its own failure class. Collapsing every rejection to `validation` flattened the connector PATCH route's 401 (stale stored credential) and 409 (missing workspace context) into a 400. - Add `unauthorized` to OrchestrationErrorCode. It is the class that 401 was already expressing on this route, and the v2 vocabulary already had UNAUTHORIZED; only the shared union was missing it. - Report a knowledge base that exists but failed to archive as failed, with the reason, rather than as not found. The copilot delete loop folded every non-not-found failure into `notFound`, telling the user it was never there. - Route copilot failures through the same message helper the HTTP surfaces use, so an unclassified fault's raw text (a driver's failed SQL) no longer reaches the agent verbatim while the UI and public APIs get the generic wording. * feat(api): expand the public v2 files surface (#6160) * feat(api): expand the public v2 files surface Adds folder support, rename/restore, move, bulk archive, share, and content replace to /api/v2/files, so managing files by API no longer stops at upload + download + archive-one. Routes are thin: auth -> parse -> perform* -> serialize. Share and content replace get their orchestration extracted first so the session routes and the public ones cannot diverge on the effective-authType resolution, the EE public-sharing gate, or the storage-quota classification. Presigned upload stays session-only: presign does an advisory quota check and the real debit happens in the separate register step, so a caller that never registers leaves unaccounted bytes with no reaper. The buffered multipart path debits inside uploadWorkspaceFile's own transaction. * fix(files): classify folder and content failures instead of 500ing them Bugbot round 1. The v2 routes map errorCode straight to a status, so every manager failure that arrived unclassified became a 500 for what is really a caller-fixable 400 or 404. - Folder manager throws OrchestrationError: missing target/folder -> not_found, reparent cycle / self-parent / restore-into-archived-workspace -> validation. - File manager does the same for the in-transaction 'File not found' paths that the earlier pass missed. - updateWorkspaceFileContent's outer catch re-wrapped everything in a bare Error, which stripped the class off StorageLimitExceededError and the new not_found alike. It now rethrows a classified failure untouched and attaches cause to the generic wrap, so asOrchestrationError can still walk the chain. - Every remaining perform* gained the asOrchestrationError branch. - renameWorkspaceFile returned the pre-update read, so the v2 PATCH reported a stale updatedAt; it now returns the timestamp it actually wrote. Docs: upload auto-suffixes a duplicate name rather than rejecting it, matching the in-app uploader. The description claimed 409 and was simply wrong. * fix(files): surface a failed upload read-back as the real error getWorkspaceFile swallows a query failure and returns null unless throwOnError is set, so a transient blip on the post-upload read reported as 'file could not be read back'. Distinguish the two: a real null after a just-committed write is an invariant break, a query failure is itself. * revert(api): drop the dedicated v2 file-folder routes File folders already live in the shared folder table as resourceType 'file' (#6045 cut them over, #6051 dropped workspace_file_folders), and the remaining file-specific folder machinery is being folded into the generic folder engine. Publishing /api/v2/files/folders/** would pin that transitional split into a public contract we'd then have to keep or break. Files stay folder-aware — folderId/folderPath on the projection, folderId on upload, and the move route — because a folder id is a folder.id and survives the unification untouched. Folder management belongs on /api/v2/folders once that surface serves resourceType 'file'; until then there is no v2 way to enumerate file folders, which is the deliberate gap. The orchestration classification fixes stay: the internal routes and the copilot file-folder tools still call those perform* functions. * fix(files): classify upload failures instead of matching their wording Bugbot round 2. uploadWorkspaceFile had the same outer-catch rewrap that updateWorkspaceFileContent did, so a blown storage quota reached the route as a bare Error and the v2 handler recovered the status by substring-matching the message. Any rewording silently demoted a 413 to a 500. - uploadWorkspaceFile rethrows a classified failure untouched and attaches cause to the generic wrap. - FileConflictError is now an OrchestrationError('conflict'), so a duplicate name classifies like every other conflict. Its 'FILE_EXISTS' discriminator had no readers and is gone; the instanceof checks elsewhere still hold. - The v2 upload handler uses v2CaughtOrchestrationError, dropping all three string matches. Also documents that bulk-archive is best-effort: unknown or already-archived ids are skipped rather than failing the call, and deletedItems is what actually happened. That asymmetry with the single-id DELETE was undocumented. * feat(api): add search, filtering, and sorting to the v2 list endpoints (#6189) * feat(api): add search, filtering, and sorting to the v2 list endpoints One convention across every v2 list, documented on lib/api/contracts/v2/shared.ts: `search` (case-insensitive substring on the resource's natural name field), `sortBy` + `sortOrder` (per-resource enum, never a free string), and enumerated resource-specific filters. Reuses the sortBy/sortOrder pair v2 logs and v2 knowledge-documents already ship rather than inventing a third dialect alongside the Logs filters and the Tables predicate grammar. Every filter and sort is pushed into SQL. GET /api/v2/files previously read the whole scope and sorted/sliced it in JS; it now goes through a new queryWorkspaceFiles that filters, orders, and bounds the page in one query. Cursors are stamped with the sort they were minted under, so replaying one under a different sort is a 400 instead of silently duplicated or skipped rows. * fix(api): validate v2 cursor key values and compare timestamps at ms precision Two review findings, fixed at the root by making a keyset key own its cursor codec instead of hand-writing a decoder per sort. Cursor key values are caller-controlled, and matching the sort stamp and key count was not enough: an unparseable timestamp or a non-numeric size reached the query as an Invalid Date or NaN and surfaced as a 500. Each key now type- checks its own value and rejects a cursor it cannot hold, which both routes render as the documented 400. Timestamp keys now order and compare on date_trunc('milliseconds', col). Postgres keeps microseconds and defaultNow() populates them, but a cursor value round-trips through a millisecond-only JS Date — comparing the raw column against the truncated value re-admitted the page's own last row, duplicating it and stalling pagination outright at a page size of one. Reachable today via workspace_files.updated_at, which insertFileMetadata leaves to defaultNow(). * feat(api): complete the v2 workflows resource with versions and CRUD (#6184) * feat(api): complete the v2 workflows resource with versions and CRUD Adds version listing/detail plus create, update, and delete to the v2 workflows surface, which previously covered only execution and deployment. - GET /api/v2/workflows/[id]/versions — cursor-paginated, newest first - GET /api/v2/workflows/[id]/versions/[version] — version + pinned state - POST /api/v2/workflows, PATCH and DELETE /api/v2/workflows/[id] All six delegate to the existing orchestration and persistence helpers; no new domain logic. * fix(api): check folder containment before lock state; reject malformed version cursors assertFolderMutable walks a folder's ancestor chain without filtering on workspace, so inspecting it before containment let a caller tell a locked folder in someone else's workspace (423) from a nonexistent one (400). Create and update now assert containment first, matching the ordering import-workflow.ts already uses. A version cursor that decodes to JSON without a numeric version filtered every row out and returned an empty page with nextCursor null, which reads as a clean end-of-list. Malformed cursors are now a 400. * refactor(api): page workflow versions in the persistence helper listWorkflowVersions read every version row and the route filtered and sliced the result in memory, so the response was bounded but the query was not. It now takes optional limit/afterVersion, turning the cursor into a real keyset query; the route asks for limit + 1 and only trims the has-more probe. Both params are optional, so the internal, v1 admin, and copilot callers are unchanged. Also restores the untouched GET handler in [id]/route.ts to its original formatting — collapsing its signature had re-indented the whole body and buried the actual additions in whitespace churn. * feat(api): expand v2 tables with stateless multipart transfers (#6188) * feat(api): expand the public v2 tables surface Adds 16 operations so a v2 caller can do what the internal surface can: rename/move/lock a table, restore it, manage saved views, run enrichment columns, look up rows, and import/export with observable job control. Extracts lib/table/orchestration/import.ts (performTableCsvImport, performCreateTableFromCsv) and lib/table/export-stream.ts from the first-party routes, then repoints those routes at them, so v1 and v2 cannot drift on what an import or export actually does. events/stream, metadata and dispatches stay internal — they are editor state, not public API. * fix(api): make v2 table PATCH all-or-nothing and name the lock in every 423 Greptile P1: PATCH applied locks, rename and move as three sequential transactions, so a folder rejected mid-request left the earlier writes persisted while the response reported failure — and the schema-changed signal was skipped, leaving open clients on stale state. Every rejectable condition now runs before the first write, and the signal fires whenever anything did land. Cursor: v2TableLockError dropped the lock kind, so async import, column run, enrichment and table mutations returned a bare LOCKED. A table has four independent locks, so the caller could not tell which to clear. * fix(api): report the lock kind on classified 423s too, not just thrown ones The previous commit named the lock only where the rejection was thrown and caught at the route boundary. Where it instead arrives as a classified `errorCode: 'locked'` outcome — delete table, delete row, update column, and the table mutations — the kind was dropped, so those 423s stayed unactionable while their neighbours improved. The orchestration results now carry `lock`, and a shared `v2TableOrchestrationError` renders both arrival paths into the same `{ code, message, details: { lock } }` body. `details` is omitted rather than sent null when the kind is unknown, so a caller branching on it sees absence instead of a phantom value. * fix(api): make async table imports observable, not just startable `POST /import-async` pointed callers at `GET /api/v2/tables/jobs` to track progress, but that endpoint filters to `type = 'export'` — imports are derived onto the table itself, one write job at a time, and exports get a separate list precisely because they are excluded from that derivation. The public Table shape omitted those derived fields, so an async import could be started and cancelled but never observed to completion, failure, or progress. That is the gap the import/export/job-control set was meant to close. Table now carries `job` — id, type, status, rowsProcessed, error, or null when idle — and the import-async docs point at the table rather than the export list. * feat(api): make v2 table PATCH state which operations landed on failure Greptile held the PR at 4/5 on the residual non-atomicity and named two acceptable resolutions: make PATCH atomic, or have the contract adopt and expose partial-success explicitly. Atomicity would mean threading one transaction through renameTable, moveTableToFolder and updateTableLocks — three shared service functions with four non-test callers including the first-party route and two copilot tools — and deferring their per-operation audits to commit time. That is a refactor of shared write paths well outside this PR. So the contract states it instead. Every rejectable condition is already pre-validated, so a failure here is a genuine fault; when one follows a successful operation the error now carries `details.applied` listing what is live. Absent when nothing applied, so its presence always means "these changes took effect despite the error". Documented on the operation. `v2ErrorForOrchestration` gained the optional `details` this needs. * fix(api): make table lock flags read-only on the public v2 surface The new PATCH /api/v2/tables/[tableId] accepted a `locks` object, gated on workspace admin plus the table-locks feature. That still lets an API key clear the guard placed there to stop it: `write` is the floor for the endpoint, and admin keys are ordinary API keys, so a lock is no longer a boundary the key cannot cross. Locks stay readable on the table resource and enforcement is unchanged (a locked verb still returns 423). Changing one is now a first-party admin action only. The v2 body is declared here rather than reusing the first-party updateTableBodySchema, which keeps its `locks` field so the UI can still toggle them. It is .strict(), so a request carrying `locks` is rejected with a 400 naming the field instead of silently succeeding without applying it. * fix(api): keep reporting applied operations when the PATCH re-read fails The composite table PATCH promises that `error.details.applied` names the operations that are live despite an error, but `applied` was scoped inside the try. A rename or move that committed and was then followed by a throw in the final re-read — or a re-read finding the table archived — returned a bare 500/404 with no details, telling the caller nothing had landed. It would then retry into a duplicate-name conflict or repeat the move. `applied` is now function-scoped so every post-write exit carries it: the 404 on a missing re-read, a thrown lock error, a classified orchestration error, and the generic 500. `v2TableLockError` gains the same `extraDetails` parameter `v2TableOrchestrationError` already had. * feat(api): add workflow group writes to the v2 tables surface v2 exposed GET /groups but none of the writes, so the public API could run an enrichment or workflow column and read its binding, but never create one. A caller could add a plain data column and trigger the machine; wiring the two together still required the UI. Adds POST/PATCH/DELETE on /api/v2/tables/[tableId]/groups. The group is the unit that fills columns — one group feeds several — so creating one creates its output columns in the same call, matching the first-party shape rather than inverting it onto the column endpoint. Four departures from the first-party body, all public-surface concerns: - group.id is optional and server-generated. The UI mints an id to render optimistically; a public caller has no such need and a client-chosen id is a collision waiting to happen. - outputColumns[].workflowGroupId is dropped from the body and stamped from the resolved group, so it cannot disagree with it. - autoRun defaults to false. First-party defaults true so a UI add fills cells immediately; here it would make one POST fan out a metered run across every existing row. - A group naming neither a workflowId (type manual) nor an enrichmentId (type enrichment) is a 400 rather than a half-specified group the route has to guess about. Also rejects an outputColumns entry no group output feeds — the two arrays are joined by column name, and the first-party client builds both from one picker so it cannot desync, but a public caller can. Workspace containment on workflowId is asserted before it is persisted, on create and on any update that re-points the group; without it a table becomes a way to invoke workflows the key cannot otherwise reach. * improvement(api): make v2 table import and export async-only Drops the three synchronous entry points: POST /tables/[tableId]/import, POST /tables/import-csv, and GET /tables/[tableId]/export. Sync import tied a write to the lifetime of an HTTP request. The body *was* the data, so it carried a 10 MB cap that Next silently truncates past — a partial import reporting success. It also had no job, so a timeout mid-write left rows in place with nothing to poll and nothing to cancel. The async path reads the file from storage instead: upload via POST /api/v2/files for a key, start with POST /import-async, watch GET /tables/[tableId] -> job, stop with POST /job/cancel. Sync export carried no such hazard, but one shape per operation beats two: with both removed the surface has exactly one way to move a table in or out, and the CLI wraps the extra calls. This also removes the last multipart handling in v2 tables. Those were the only routes bypassing parseRequest — form fields were parsed by hand against separate form schemas, outside the contract system every other v2 write goes through. Create-a-table-from-CSV is now two calls: POST /tables, then /import-async with createColumns. csvImportModeSchema is append|replace, so there is no single-call create. Route baseline 1064 -> 1061. * docs(api): correct the import-async note about upload size limits The docstring claimed there is no synchronous upload endpoint and so no request-body size cliff. Both are wrong: POST /api/v2/files is a synchronous multipart upload with a 100 MB cap, and it is the only v2 upload path (presigned is deliberately absent). What async-only actually bought: the cap went 10 MB -> 100 MB, it fails on an explicit size check and a bounded body read rather than a proxy cap that silently truncates, authorization completes before any body is buffered, and the table write is a job that can be watched and cancelled. * feat(api): unify file and table transfers * improvement(api): make multipart transfers stateless * fix(api): make table import completion retries idempotent * feat(v2-tables): paginate the table list `GET /api/v2/tables` returned every table in the workspace in one response — it used the cursor envelope but hardcoded `nextCursor: null`, and had no `limit`. That was defensible when tables were only created through the UI; `POST /api/v2/tables` is public now, so a script can create them in bulk and the list has no way to ask for less. Adds `queryTables` alongside `listTables` rather than changing it, so the internal callers that genuinely want the whole scope are untouched — the same split `queryWorkspaceFiles` / `listWorkspaceFiles` already uses. Filter, order and slice all run in the query, so a `search` never costs a full-workspace read. A cursor whose values don't bind raises a validation error instead of being coerced to "no filter", which would have silently served page 1 under a resumed cursor. The keyset closes on `id` so a page boundary inside a run of equal names or timestamps stays stable. The shared `LimitQuery` doc component said "Maximum rows to return"; it now serves the table list too, so the wording is resource-neutral. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat(api): add multipart knowledge document uploads * fix(api): keep usage admission at knowledge upload session creation * feat(knowledge): wire knowledge base uploads to multipart sessions * fix(knowledge): refuse to abort an upload once a document is bound * fix(uploads): prevent multipart cleanup races * Unify file creation and signed upload sessions (#6264) * feat(uploads): unify signed upload sessions * fix(uploads): preserve attachment storage semantics * feat(files): add authored file creation * fix(uploads): omit hoisted S3 metadata headers * feat(api): add file metadata endpoint * improvement(api): scope folders to resource paths (#6284) * improvement(api): scope folders to resource paths * fix(files): serialize folder resolution with uploads * fix(files): release folder lock before upload setup * fix(api): normalize folder paths and unblock resource mutations * fix(api): make resource cleanup and metadata consistent * improvement(uploads): persist multipart sessions in postgres * fix(db): store table row trigger timestamps in UTC * improvement(api): default folder deletion to non-recursive * fix(billing): unify chat usage source * improvement(logs): expose trace spans on log detail * fix(logs): parse list trace spans * improvement(api): replace workflow jobs with execution resources (#6294) * improvement(api): replace workflow jobs with execution resources * fix(api): preserve legacy jobs while preferring v2 executions * fix(api): make execution polling resume-aware * fix(ui): hide async examples for public workflows * fix(api): bridge resume queue visibility lag * feat(api): add v2 workflow resume endpoint * fix(api): project pending resume attempts * fix(api): prefer terminal logs over stale resumes * improvement(api): unify v2 resource query layers (#6319) * improvement(api): unify v2 resource query layers * fix(api): address v2 review findings * fix(api): preserve cancelled queue status * fix(api): guard cancelled job transitions * fix(api): close v2 resume and log gaps * feat(api): rename v2 executions to runs * feat(api): split credentials and secrets * feat(api): add workspace metadata and email attribution * improvement(api): consolidate public v2 route handling * improvement(files): centralize operations across APIs and Copilot (#6392) * improvement(files): unify rename authorization * chore(skills): add file operation migration guide * improvement(files): consolidate file operation authorization * improvement(files): extract shared operation foundation * improvement(api): simplify internal route declarations * improvement(files): centralize application authorization * refactor(api): share workspace file name validation * refactor(files): centralize copilot application calls * docs(skills): generalize application operation migration * improvement(api): centralize remaining v2 resource operations (#6412) * improvement(api): centralize v2 resource operations * fix(api): preserve custom tool conflict errors * improvement(api): migrate policy-sensitive v2 reads (#6410) * improvement(workflows): centralize v2 application operations (#6411) * refactor(api): migrate v2 knowledge operations (#6413) * refactor(api): migrate v2 knowledge operations * fix(knowledge): fail upload completion on dispatch errors * fix(knowledge): preserve upload retry and VFS errors * improvement(tables): centralize v2 application operations (#6414) * improvement(tables): centralize v2 application operations * fix(tables): preserve run validation and signals * feat(auth): add scoped internal executor delegation (#6459) * feat(auth): add scoped internal executor delegation * fix(auth): derive delegation lifetime from one timestamp * Include share status in file metadata * feat(auth): centralize delegated identity policy (#6462) * improvement(copilot): consolidate application adapters (#6450) * improvement(api): harden application route boundaries (#6451) * improvement(api): harden application route boundaries * fix(folders): reject creates at workspace cap * fix(knowledge): enforce trusted workspace scope (#6452) * fix(knowledge): enforce trusted workspace scope * refactor(knowledge): declare v2 body lifecycle * finish knowledge application migration * refactor(knowledge): compose copilot batch commands * fix(knowledge): parse connector query flags * fix(knowledge): finalize partial batch effects * fix(knowledge): align merged application boundaries * fix(knowledge): close application boundary review gaps * style(knowledge): satisfy branch biome checks * fix(knowledge): page connector documents in editor * refactor: enforce Copilot table application boundary (#6453) * refactor: enforce copilot table application boundary * fix(tables): finish application boundary migration * fix(tables): restore scoped copilot imports * fix(tables): compose copilot commands atomically * fix(tables): preserve workflow group scheduling * fix(tables): complete fixed copilot composition * fix(tables): reject enrichment output mutation * fix(tables): complete authorized application boundary * fix(workflows): migrate Copilot application boundary (#6455) * fix(workflows): migrate Copilot application boundary * fix(workflows): finish delegated application migration * fix(workflows): encode VFS folder aliases * fix(workflows): close application composition gaps * fix(workflows): preserve VFS validation errors * fix(workflows): complete application boundary migration * test(workflows): format canonical binding coverage * fix(workflows): scope executor metadata reads * fix(workflows): bind executor metadata targets * improvement(skills): align application operation guidance (#6532) * feat(api): expose v2 resource owners * fix(api): distinguish visible resource authorization failures (#6537) * feat(api): generate v2 OpenAPI from contracts (#6509) * feat(api): generate v2 OpenAPI from contracts * fix(api): preserve string boolean wire defaults * fix(api): document file download headers * fix(docs): use TypeScript CLI with Next.js * fix(docs): avoid client-rendered theme script * fix(api): document departed audit default * feat(api): replace legacy core docs with v2 * feat(api): generate v2 OpenAPI from contracts * feat(api): refine generated v2 OpenAPI docs * fix(docs): align localized v2 execution examples * fix(ci): restore Helm diff and sync audit mock * fix CI regressions after staging merge --------- Co-authored-by: Waleed <walif6@gmail.com> Co-authored-by: Theodore Li <theodoreqili@gmail.com> Co-authored-by: Siddharth Ganesan <33737564+Sg312@users.noreply.github.com> Co-authored-by: Theodore Li <theo@sim.ai> Co-authored-by: Claude Fable 5 <noreply@anthropic.com>	2026-08-11 05:45:25 -04:00
Waleed	dcaa118752	improvement(docs): restructure sidebar, align chrome, rename Mothership to Chat (#6296) ... Sidebar: 11 separator groups become 5, with each module a collapsible folder that auto-opens on the active page. 61 always-visible rows drop to 16. Groups mirror the app's own nav (Chats/Workspace/Workflows) rather than inventing a taxonomy; Enterprise and Self-Hosting are hoisted out of Platform. Chrome: register the `hover-hover` variant, without which every @sim/emcn hover state silently compiled to nothing; restore the sidebar's Geist font stack; add 11 emcn tokens that were falling back to currentColor; adopt the named type scale; align row geometry, hover tokens and group labels with the app. Rename: mothership/ -> chat/ with redirects for the old URLs. Asset paths, the @mothership.sim.ai domain and the `mothership` log-trigger enum value are deliberately left alone -- they are CDN objects, a real domain, and a live product value. Also removes the page-type badge, drops the "Next" heading from the ToC, and lets FAQ rows open independently so expanding one no longer shifts the page.	2026-08-05 13:49:17 -07:00
Waleed	fc25cfb3c8	fix(docs): fix Core Web Vitals regressions on docs.sim.ai (#5630) ... * fix(docs): fix Core Web Vitals regressions on docs.sim.ai Empirically measured under real trace-based (devtools) CPU/network throttling against the live site: mobile Performance 59, LCP 9.2s (TTFB 745ms + 8.4s element render delay). - sidebar-components.tsx / [lang]/layout.tsx: the docs sidebar renders every page in the doc tree as a link at once. Next's default viewport-prefetch fired an RSC payload fetch for every one of them on initial load - dozens of concurrent requests competing with the page's own content for bandwidth. Wired fumadocs' documented `sidebar.prefetch` option through to the custom SidebarItem/SidebarFolder components (which were bypassing it entirely, using next/link directly with no prefetch prop) via the `useSidebar()` context hook. - video.tsx: `autoPlay` forces browsers to fetch the full video file immediately on mount regardless of `preload`. Gated actual src loading behind an IntersectionObserver so a page with several of these doesn't pull down every video up front (5MB across 3 requests, in this case). Single shared component - fixes every doc page that embeds one. - proxy.ts: the i18n middleware matcher excluded favicon/robots.txt/etc but not `icon.svg`, so every request for it got routed through i18n negotiation instead of served as a static file, 404ing in production. - next.config.ts: enable productionBrowserSourceMaps - safe since this repo's source is already fully public, real debuggability benefit, zero performance cost. - shiki 4.0.0 -> 4.3.1 (verified: syntax highlighting still renders correctly). Attempted a coordinated fumadocs-core/ui/mdx/openapi upgrade to latest; fumadocs-openapi's v11 factory function became client-only (breaking change beyond its declared peer deps, requiring a component-boundary restructure), so only the safe, verified, docs-exclusive bumps (fumadocs-core/ui/mdx, shiki) are included here - the openapi major bump needs its own dedicated migration PR. Verified via a real production build (dummy env, all 3974 pages including API reference render/build cleanly) and a clean (non-stale) local server: Performance 59 -> 71 measured under real devtools throttling, RSC prefetch requests 63 -> 11, video requests/bytes 3/5MB -> 0. A pre-existing React hydration warning (#418) was found and confirmed present on live production before any of these changes, unrelated to this diff - documented, not blocking. * fix(docs): fall back to eager video loading without IntersectionObserver The lazy-load gate from the previous commit threw before isInView could ever become true in environments lacking IntersectionObserver (older browsers, some embedded webviews), leaving videos permanently source-less instead of falling back to eager loading. * chore(docs): drop non-TSDoc inline comments Repo convention is TSDoc-only, no plain // comments. * fix(docs): accessibility and SEO defects across the docs app Audited with parallel subagents against the accessibility and SEO skill checklists, each fix verified by reading the actual code (not assumed): Accessibility: - lightbox.tsx: focus was never captured/restored on close, and Tab escaped the modal to the page behind it (no focus trap on the single focusable element) - heading.tsx: the per-heading copy-link icon only appeared on hover, invisible to keyboard-only navigation (added peer-focus-visible) - navbar.tsx: active nav tab had no aria-current - response-section.tsx: the status-code dropdown had no aria-haspopup/aria-expanded/role, and no Escape-to-close - workflow-preview.tsx: same focus-trap gap as lightbox.tsx on the expanded-canvas modal SEO: - page.tsx: generateMetadata's hreflang/canonical URLs used a naive String.replace to strip the locale prefix, which also matched "/en" inside unrelated slugs (platform/enterprise, integrations/enrich, platform/self-hosting/environment-variables), corrupting those pages' canonical and alternate-language URLs. Replaced with a prefix-only strip. - structured-data.tsx: the SoftwareApplication JSON-LD block compared url === baseUrl (no trailing slash) against the homepage's actual url (always has a trailing slash), so the condition was always false and this structured data never rendered anywhere, including the homepage. - structured-data.tsx: "Mothership" in the indexed featureList violated the constitution's required language (the agent is "Sim", the surface is "Chat") - this ships in JSON-LD search engines parse. * fix(docs): defer the Ask Sim chat widget's heavy deps until opened The chat panel (useChat from @ai-sdk/react, Streamdown + its CSS) was mounted unconditionally in the root layout on every single page, so its full weight loaded and executed even though the widget starts closed on every page view. Traced via the LCP breakdown insight under real devtools CPU/network throttling: the LCP text element (the intro paragraph) had a ~8s element render delay despite a ~13ms TTFB, and bootup-time attributed ~4.3s of scripting time to a single chunk containing React/ReactDOM's own runtime plus this widget's eagerly-bundled dependencies. Split into a lightweight ask-ai.tsx (just the toggle button + open state) and ask-ai-panel.tsx (the actual chat UI, useChat, Streamdown), loaded via next/dynamic(..., { ssr: false }) only when the user opens the widget. Verified: the panel's chunk now has zero network requests on initial page load. Measured (mobile, devtools throttling, /introduction): - Performance: 69 -> 75 - LCP: 8.0s -> 6.4s - TBT: 260ms -> 130ms The remaining ~6.4s LCP delay traces to the same shared chunk, now identified as core React/ReactDOM hydration cost for this page's sidebar/TOC/breadcrumb tree rather than an isolated bug - a real, larger initiative (hydration architecture, not a surgical fix), documented here rather than rushed. * fix(docs): preserve Ask Sim chat state across close/reopen The panel split unmounted AskAIPanel entirely on close, discarding useChat's message state - reopening always started an empty conversation, unlike the original single-component layout where useChat lived in a component that never unmounted. Fixed by keeping the panel mounted (via a hasOpened flag that never resets) once first opened, and having the panel itself return null when closed rather than being conditionally removed from the tree by its parent - hooks still run every render, so useChat's state persists across visibility toggles. The dynamic import still only fires on the first open, so the initial-load win is unchanged. Verified via a real click-through (open, type, close, reopen): input persists correctly, and the panel chunk still has zero network requests on initial page load. Performance unchanged at 75. * chore(docs): lint fixes (import order, formatting) * fix(docs): fill the Ask Sim UI gap while the panel chunk loads handleOpen set open=true synchronously, hiding the trigger button before the dynamically imported panel had a chance to render anything (next/dynamic renders null by default with no loading option) - on a slow connection neither the button nor the panel was visible. Added a loading fallback in the same fixed position so there's no gap between the button disappearing and the real panel appearing.	2026-07-13 09:10:33 -07:00
Waleed	7545391cb3	feat(docs): render workflow previews with the shared editor renderer (#5277) ... * chore(workflow-renderer): declare @sim/emcn dep + wire the package into docs Adds the missing @sim/emcn peer/dev dependency to @sim/workflow-renderer (it imports @sim/emcn in every View but resolved only via workspace hoisting). Wires apps/docs to consume @sim/workflow-renderer (dependency, transpilePackages, Tailwind @source) and adds remark-breaks (pulled transitively via the barrel's NoteBlockView export) — mirroring the @sim/emcn integration. Foundation for migrating the docs workflow-preview fork onto the shared Views. Build resolves the package/@source/remark-breaks cleanly. * feat(docs): render loop/parallel containers with the shared SubflowNodeView Replaces the forked PreviewContainerNode with a thin DocsContainerNode that maps the static preview data to SubflowNodeView's read-only (isPreview) props — no stores or hooks. Adds the block size to the preview node data so the view can size itself, and corrects the parallel example's start-edge handle id to 'parallel-start-source' (the view derives the handle id from kind). Deletes preview-container-node.tsx. Container colors/icons are now owned by the shared view (loop=blue, parallel=yellow). * feat(docs): render block nodes with the shared WorkflowBlockView Replaces the forked PreviewBlockNode with a thin DocsBlockNode that maps the static preview data to WorkflowBlockView's props — store-free, builds the subblock rows (condition/router Context+routes/default + tools + error) via SubBlockRowView, strips branch-id prefixes so the view's regenerated handle ids match, remaps router->router_v2, and keeps the framer-motion dim/stagger wrapper. Promotes resolveIcon into block-icons.tsx, adds the --workflow-edge token to docs global.css, deletes preview-block-node.tsx. The canvas diagrams now render with the real editor's view. * refactor(workflow-renderer): make editor-only WorkflowBlockView props optional The child-deploy, schedule, and webhook badge props (and their callbacks) only matter in the editor. Mark them optional and optional-chain the three callbacks so read-only consumers (docs, academy) can omit the whole group instead of passing ~18 explicit off-values. The editor still passes them, so its behavior is byte-identical (verified: apps/sim type-check clean). DocsBlockNode drops the off-props. * feat(docs): replace how-it-runs static diagrams with live WorkflowPreview Swaps the four static PNGs on the how-it-runs page for live, app-styled WorkflowPreview diagrams (concurrency, combination, condition+router branching, error path). Adds the four example workflows and renders error-port edges red to match the editor. The English page only; the translated execution/basics pages keep the PNGs. * refactor(workflow-renderer): the view owns condition/router/error rows Both the editor container and the docs adapter hand-built the condition/router/error summary rows in an order that had to stay in lockstep with the view's absolute handle-offset math — a three-way coupling with nothing enforcing it. The view now renders those rows itself from the conditionRows/routerRows it already receives (plus a routerContextValue prop for the router's Context row), so row order and handle geometry live together in one place. Both containers pass only data and their non-branch rows. Editor is byte-identical: getDisplayValue moves to where conditionRows/routerRows are built; the no-subBlock SubBlockRow path is already an exact SubBlockRowView(title, value) passthrough; the error row stays gated on shouldShowDefaultHandles. Verified apps/sim type-check clean. Docs now also renders the error row on condition/router blocks, which the real editor already did (shouldShowDefaultHandles is true for them) — an alignment fix. * refactor(docs): drop the parallel --wp-* token layer for the app/emcn tokens The workflow-preview ran a 25-token --wp-* mirror (22 were pure aliases of app tokens docs already defines) plus a .wp-scope wrapper class. Replaces every var(--wp-X) with its canonical app/emcn token (--wp-edge->--workflow-edge, --wp-highlight->--brand-secondary, badges->--badge-*, etc.), adds the one missing token (--divider), and deletes the .wp-scope blocks + class. Visually identical (aliases resolve to the same values); the preview now inherits the same design tokens as the shared views and the rest of the app instead of a hand-rolled parallel set. * refactor(docs): adopt emcn Badge + dedup resolveIcon in workflow-preview output-bundle's hand-rolled type badge (BADGE_COLORS + a styled span) becomes the emcn Badge (its green/blue/orange/purple/gray variants use the identical --badge-* tokens). resolveIcon, which had three copies, is now imported once from block-icons by output-bundle and block-inspector. * refactor(docs): rebuild the preview inspector on emcn chip primitives The lightbox inspector was a hand-rolled facsimile (raw divs + a CONTROL class string + inline dashed borders). It now composes from the same @sim/emcn primitives the live editor's sub-block controls wrap — ChipSelect/ChipInput/ChipTextarea(viewOnly)/ChipSwitch/ChipTag/FieldDivider/Label — so it reads as the real editor panel, fed example data (read-only, full opacity via readOnly/viewOnly, not greyed). Slider stays minimal (no emcn equivalent) but on app tokens. Props API and embedded/standalone modes unchanged. * refactor(docs): render the block-reference hero through the shared View Retires the hand-rolled BlockCard (a parallel reimplementation of WorkflowBlockView) and the BlockDisplaySpec data model. Each block hero is now a single-block PreviewWorkflow (block-display-workflows.ts) rendered through the same toReactFlowElements -> DocsBlockNode -> WorkflowBlockView pipeline as the diagrams, mounted in a minimal fitView ReactFlow (maxZoom 1.3, no canvas chrome). A single block can no longer drift from the canvas. * fix(docs): define sim's type scale + align the preview inspector to the editor Docs Tailwind v4 never defined sim's custom font sizes (text-small/caption/md/micro), so emcn components (Label, Badge, the shared views) fell back to inherited sizes — the inspector labels rendered huge. Adds the type scale to the docs @theme. Also aligns the inspector header to the real editor panel (surface-4 bar, size-[18px] rounded-sm icon, text-sm name) and removes the Connections section (and its now-dead prop/wiring). * fix(docs): inspector shows the full field list + dragged positions persist Inspector: shows the block type's full field list (from the reference data) with the example's values overlaid, so it reads like the editor panel instead of only the canvas summary rows. Drag: selecting another block no longer relayouts the canvas — node positions the viewer dragged are preserved across highlight/selection changes (only a different workflow relayouts). * feat(docs): highlight <> references + env vars; hide Ask AI over the lightbox; respace blocks Inspector text fields render the value with <...> block references and {{...}} environment variables highlighted in brand-secondary (a lean read-only port of the editor's formatDisplayText), in the canonical chip field chrome. The floating Ask AI widget is hidden while a preview lightbox is open. Plus the example-data respacing so the editor-faithful Error row no longer makes stacked blocks overlap. * fix(docs): make per-type field templates match the real block registry Audited every block type's field list (the source the inspector + block-reference heroes render) against apps/sim/blocks/blocks/*. Corrected drift to the registry's default-visible fields, titles, and order: agent gains Temperature; router gains Model; wait gains Async; schedule rewritten (default is Daily, not minutes); webhook_trigger expanded to its real default-visible set; human_in_the_loop notification title fixed. Provider-credential and advanced-mode fields stay hidden, matching the editor. Canvas diagrams keep their clean curated rows; the inspector now shows the full, real field list per the chosen clean-canvas/full-inspector split. * improvement(docs): taller default preview height so respaced diagrams aren't shrunk Bumps the default WorkflowPreview height 260->300 (the respaced, editor-faithful blocks are taller, so fitView was shrinking diagrams that relied on the default). The tall how-it-runs routing diagram gets 400. * improvement(docs): zoomable inline preview + taller default + themed controls The inline preview is now zoomable outside the lightbox: adds react-flow zoom/fit Controls (themed to the dark canvas chrome) and enables pinch-zoom, while keeping scroll-zoom off so the page still scrolls over the diagram. Pan-drag and click-block-to-inspect already worked. Default height 300->340. * improvement(docs): click canvas to expand; click empty lightbox to deselect Clicking the inline preview canvas opens the full lightbox; clicking empty space in the lightbox clears the selection, matching the real editor. * improvement(docs): reveal inline zoom controls on hover only The always-visible zoom controls felt heavy on the inline preview; they now fade in on hover (matching the expand button) and stay visible in the lightbox. * improvement(docs): drop zoom controls on the inline preview Inline preview keeps pinch-zoom, pan, drag, and click-to-expand; zoom buttons stay in the lightbox only. * improvement(docs): remove zoom controls from the lightbox too Both previews zoom via scroll/pinch and pan via drag; no on-canvas zoom buttons. Drops the Controls import and its theming CSS. * improvement(docs): match the real canvas — flat background + editor edge geometry Closes the last faithfulness gaps the audit found: removes the dot grid (the real editor hides its background — flat bg), aligns PreviewEdge to the editor's smoothstep math (borderRadius 8, offset 30) and 2px stroke (default + error edges), the selection ring to 1.75px, and minZoom to 0.1. Structural parity (blocks/handles/containers/colors/tokens) was already shared. Kept PreviewEdge rather than swapping to WorkflowEdgeView, which would clobber the docs-only highlight/dim/animate for no visual gain. * improvement(docs): rebrand the docs assistant as 'Ask Sim', styled like the real chat input Renames the floating assistant from 'Ask AI' to 'Ask Sim' (matching the platform's voice — you talk to Sim) and restyles the composer to mirror the home chat input: a rounded-2xl bordered field with the toolbar inside, and the same 28px circular send/stop button (the home's exact active/disabled colors + white/black arrow). Updates the lightbox hide-selector to the new label. * improvement(docs): match Ask Sim message styling to the mothership chat Aligns the user bubble (rounded-[16px] surface-5, text-base/primary, leading-23, max-w-85%) and the assistant markdown (text-base, 600 headings/strong, text-primary dashed-underline links, surface-5 code blocks) to the real mothership chat's user-message + chat-content treatment, instead of the prior generic text-sm rendering. The composer already mirrors the home user-input (rounded-2xl field + 28px circular send button). * improvement(docs): compact single-row Ask Sim composer The two-row layout left a tall dead gap (the docs widget has no toolbar buttons to fill the second row). The composer is now a single row — textarea with the circular send button inline — so it sits at the natural input height. * fix(docs): pass the router Context value to the shared view DocsBlockNode never set routerContextValue, so the view (which renders the router's Context row from that prop, not from rows) showed a blank Context even when the preview data authored a value like <start.input>. Extract it from the block's Context row and pass it through. * fix(docs): don't apply a block-type field template that doesn't match the block inspectorFieldsFor keyed the full field template purely off block.type, but some types are reused across roles (a table action block vs the table trigger, a webhook trigger vs the webhook action), so the wrong template was applied. Only use the template when the block's authored rows are actually a subset of it; otherwise fall back to the block's own rows. * fix(docs): connect preview edges to subflow container handles toReactFlowElements hardcoded targetHandle to 'target' and defaulted source handles to 'source', but Loop/Parallel containers (SubflowNodeView) expose a 'loop-end-source'/'parallel-end-source' output handle and a left input handle with no id. Edges into and out of containers therefore failed to connect. Resolve each edge end to the block's real handle based on whether it's a container. * fix(docs): don't expand the inspector template for blocks with no rows block.rows.every(...) is vacuously true for an empty rows array, so a block defined only by branches (e.g. a router in ROUTING_WORKFLOW) inherited the type template's invented field defaults. Require non-empty authored rows before applying the template. * fix(docs): render blank branch/router-context values as '-' like the editor The editor maps condition/router branch values and the router Context through getDisplayValue, which renders '-' for a blank value. DocsBlockNode mapped them to an empty string, so else branches and unset routes looked blank instead of matching the editor. Mirror getDisplayValue's empty-value handling. * fix(docs): show '-' for blank inspector branch values, matching the canvas inspectorFieldsFor passed raw branch.value into the lightbox branch fields, so an unset else route read blank in the inspector while DocsBlockNode (and the editor's getDisplayValue) render '-' on the canvas. Normalize the same way; drop the now-redundant placeholder.	2026-06-29 20:27:33 -07:00
Waleed	bcf6a804f9	improvement(emcn): extract design system into shared @sim/emcn package (#5257) ... Moves apps/sim/components/emcn into a shared @sim/emcn package consumed directly by apps/sim and apps/docs. cn/keyboard/use-copy-to-clipboard move into the package; all imports become direct @sim/emcn (icons via @sim/emcn/icons, CSS via file path). ChipModal email validation is now prop-driven (quickValidateEmail stays in apps/sim, injected via validate). Docs drops its local chip/chip-dropdown/dropdown-menu mirrors and consumes @sim/emcn.	2026-06-28 22:50:48 -07:00
Will ChenandClaude Opus 4.8	bc55fc3b50	improvement(docs): builder-first IA reorganization of the English docs (#4896) ... * docs: reorganize into topic/ontology IA with a builder-first rewrite Restructure the English docs from internal product categories into a topic-based information architecture, and rewrite the conceptual pages to install a mental model first rather than enumerate features. Structure & navigation - Reorder the sidebar to follow how someone builds: Get Started -> Workflows -> Tables -> Files -> Knowledge Bases -> Logs -> Building agents -> Mothership -> Workspaces -> Platform -> Reference. - Demote the generated blocks/tools/triggers catalogs to a Reference section at the bottom. - Break up the monolithic execution/ folder into deployment/ and logs-debugging/; collapse connections/* and variables/* into single pages under workflows/. - Rename capabilities/ to building-agents/; relabel the integration catalog as "Integrations". Remove deprecated copilot and form deployment. Redirects added in next.config.ts for every moved URL. Conceptual rewrites - Workflows core (index, how-it-runs, data-flow, connections, variables): one mental model, one running example, terser prose. - New building-agents overview distinguishes an agent (a workflow you build) from an Agent block (one reasoning step), plus a "choosing what to use" guide. - Concept-trim passes on Knowledge Base, Tables, Blocks, Triggers overviews; new task pages for KB, Tables, and Files. - New code-verified Alerts page. Infrastructure - pageType frontmatter (concept/guide/reference) + badge render. - WorkflowPreview / OutputBundle components to embed real, app-styled workflow diagrams (adds framer-motion + reactflow to apps/docs). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * feat(docs): spec-driven BlockPreview for block reference heroes Replace the static screenshot hero on each block reference page with a <BlockPreview> that renders the block exactly as the builder canvas shows it — header icon, sub-block rows, and branch/error handles — from a hand-authored display spec. Static and non-interactive (no ReactFlow), so it can't be panned or dragged, and self-updating to edit. - block-display-specs.ts: one editable spec per block (rows, branches, handles) - block-preview.tsx: static scaled card renderer with decorative handles - block-icons.tsx: brand glyphs for the core block types; icons.tsx adds WaitIcon - 14 block + 3 trigger pages swapped from <Image> to <BlockPreview> Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs): correct stale navigation and removed-feature references Audited the docs against the product changelog (GitHub releases / staging git history) for content that misleads readers — features that moved, were renamed, or removed — rather than cosmetic drift. Fixes: - Skills: no longer a Settings tab. It was promoted to its own workspace page (#4354), so "Settings → Skills under the Tools section" sent readers to a tab that no longer exists. (skills/index.mdx) - Env vars: the workspace tab is "Secrets", not "Environment Variables" (credentials→secrets rename, #4364). (quick-reference/index.mdx) - Mothership FAQ pointed to "Settings → Credentials" for integration connections; integrations moved to their own page and there is no Credentials tab. (mothership/tasks.mdx) - Vision block was retired (#4684); a tip still named it. Reworded to "an Agent using a vision-capable model". (files/passing-files.mdx) - Getting-started FAQ told new users to "use the Copilot feature" to build in natural language — that surface is Mothership. (getting-started) - Removed the dead "Mod+Y → Go to templates" shortcut; the templates gallery was removed (#4354). (keyboard-shortcuts) Note: MCP "tools" (Settings → Tools, for consuming) and MCP "servers" (Settings → System, for exposing) are distinct surfaces — both doc references are correct and were intentionally left as-is. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs): repair broken /docs-prefixed enterprise links The enterprise overview linked to /docs/enterprise/* (access-control, sso, whitelabeling, audit-logs, data-retention, data-drains), but the docs site is served at root — those 6 links 404'd. Now root-relative /enterprise/*. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs): refresh stale workflow-preview example blocks The /workflows diagram blocks are hand-authored (separate from the spec-driven BlockPreview heroes) and had drifted from the real UI: - Agent color purple #6f3dfa -> green #33C482 (the var(--brand) rebrand) - Model gpt-4o -> claude-sonnet-4-6 (current default) - "Prompt" row -> "Messages" (the actual agent sub-block) - Start color #34B5FF -> #2FB3FF (real starter bgColor) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs): align BlockPreview input/output handles to the card edge The header (input/output) handles are positioned relative to the card and used a -16px offset, so they floated 8px past the edge. Row/error handles are -16px relative to a row that's already inset 8px by content padding, so they sit correctly. Header handles are now -8px, so every handle sticks out the same 8px and hugs the block edge. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Agent reference to match the current block The page documented the old UI (System/User Prompt, no Files or Skills, Memory taught as a separate block — contradicting its own FAQ). Rewritten to the real sub-blocks (Messages, Model, Files, Tools, Skills, Memory, Response Format) in the builder voice of the workflows exemplars: oriented opening, agent vs Agent-block callout, outputs table, a live WorkflowPreview example, FAQ kept and corrected (tool control "Force", not "Required"). pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite API reference to match the current block Tightened to the builder voice and the real config (URL, Method, Query Params, Headers, Body + Advanced timeout/retries/backoff). Dropped the off-topic "Dynamic URL Construction" / "Response Validation" sections (those are Function-block techniques, not API config). Outputs table, FAQ kept. The example is now a live WorkflowPreview (new API_FETCH_WORKFLOW in examples.ts, exported via the barrel). pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Condition reference to match the current block Tightened to the builder voice: oriented opening (branches on boolean expressions, no model call, vs Router), the real branch model (if / else if / else, checked top to bottom), connection-tag expression examples, an error-path callout, outputs table, and a live branching WorkflowPreview example (CONDITION_ROUTE_WORKFLOW). FAQ kept. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): restore Best Practices + multi-example workflows on Condition Recalibration: reference pages keep genuine substance (Best Practices, every distinct example), cutting only redundancy and verbose register. Restores the Best Practices section and turns the three use cases into three rendered WorkflowPreview examples (route by priority, moderate content, branch onboarding). Adds CONDITION_MODERATE_WORKFLOW and CONDITION_ONBOARD_WORKFLOW. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): restore Best Practices on Agent reference Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): restore Best Practices on API reference Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Function reference to match the current block Fixed the verbose register and dropped the duplicated outputs section + the stale Python screenshot/TODO, while keeping the real substance: JS vs Python (local vs E2B sandbox), the large-inputs sim.files/sim.values helpers, the worked loyalty-score example, and Best Practices. The use cases are now two rendered WorkflowPreview examples (reshape an API response, validate input). Adds FUNCTION_RESHAPE_WORKFLOW and FUNCTION_VALIDATE_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Router reference to match the current block Cleaned the register, generalized the drifting model list, and folded the Router-vs-Condition guidance into a callout. Kept the substance (routes as output ports, NO_MATCH error path, all seven outputs, Best Practices, FAQ). The three same-shape use cases collapse to one rendered triage WorkflowPreview (ROUTER_TRIAGE_WORKFLOW), which the prose notes stands for the pattern. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): restore the classify and lead-qual examples on Router I wrongly folded two distinct Router scenarios into a note. Restored all three as their own rendered WorkflowPreview examples: triage a support ticket, classify feedback (to child workflows), qualify a lead (sales vs self-serve). Adds ROUTER_CLASSIFY_WORKFLOW and ROUTER_LEAD_WORKFLOW. (Also exports RESPONSE_API_WORKFLOW for the next page.) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Response reference to match the current block Cleaned the register and broadened "Variable References" to connection tags (any output, not just workflow variables). Kept the substance: exit-point semantics, Builder/Editor mode, status codes, headers, the parallel-branch warning, Best Practices, FAQ. All three use cases are now rendered WorkflowPreview examples (API endpoint, webhook ack, status-per-branch). Adds RESPONSE_API/WEBHOOK/ERROR_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Variables reference to match the current block Cleaned the register, corrected the outputs (each assignment is also exposed as <variables.name>, not "no outputs"), and kept the substance: assignments reference earlier outputs and current values, global <variable.name> access, Best Practices, FAQ. Two use cases now render as WorkflowPreview examples (count retries, hold config). Adds VARIABLES_RETRY_WORKFLOW and VARIABLES_CONFIG_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Wait reference to match the current block Corrected a real staleness: the block now has an Async mode that suspends the run for minutes/hours/days (not a hard 10-minute cap), plus a resumeAt output. Documents Wait Amount / Unit / Async, the sync-vs-async distinction, all three outputs, Best Practices, and updated FAQ. Two rendered WorkflowPreview examples (space out API calls, delayed follow-up). Adds WAIT_RATELIMIT_WORKFLOW and WAIT_FOLLOWUP_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): polish Credential reference (frontmatter, fold redundant tabs) The page was already accurate to the block (Select/List operations, the outputs tabs, the wiring steps). Light touch only: added description + pageType, made the header consistent, and folded the two identical Gmail/Slack "how to wire" tabs into one line. Examples stay as labeled flows + the List/ForEach screenshot, since they use integration blocks and a Loop the WorkflowPreview can't render. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): render the shared-credential example + icon fallback for integrations Addressing the gap: WorkflowPreview block nodes now fall back to the integration icon map, so diagrams can show Gmail/Drive/Slack/etc. with their real glyphs, not just core blocks. Renders the Credential "share one account across blocks" example as a WorkflowPreview (CREDENTIAL_SHARE_WORKFLOW). The multi-account and List+ForEach examples stay as labeled flows + screenshot (the latter uses a Loop container the preview can't render). Also exports EVALUATOR_GATE_WORKFLOW for the next page. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rewrite Evaluator reference to match the current block Cleaned the register, generalized the drifting model list, and documented the per-metric outputs (<evaluator.metricname>), which the page omitted. Kept the substance (metrics with name/description/range, structured-output guarantee, Best Practices, FAQ). The quality-gate example renders as a WorkflowPreview; the same shape covers the parallel-variations and support-QC patterns, noted in prose. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): render the Credential route-by-logic example too The icon fallback unblocked it: the "route to a different account by logic" example now renders as a WorkflowPreview (CREDENTIAL_ROUTE_WORKFLOW), a Condition selecting a production vs staging credential. The List + ForEach example stays a screenshot because it nests blocks in a Loop container the flat WorkflowPreview can't represent. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): render Guardrails examples + light accuracy pass Kept the full substance (four validation types, PII entity/language detail, the PII screenshot and video, outputs, Best Practices, FAQ). Light fixes: frontmatter, and generalized the drifting model names (GPT-4o / Claude 3.7) to "a strong reasoning model" with the current default. The three use cases now render as WorkflowPreview examples (validate JSON, check grounding, block PII). Adds GUARDRAILS_JSON/HALLUCINATION/PII_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): render Human-in-the-Loop examples + frontmatter Kept all the substance (Display Data, Notification, Resume Form, the Approval Methods and API Execute Behavior tabs, outputs, the paused/resume example). Added frontmatter and rendered the use cases as WorkflowPreview examples (approve before publish, two-stage approval, verify extracted data); Quality Control folds into the approval note as the same approve-then-act shape. Adds HITL_APPROVAL/MULTISTAGE/VALIDATE_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): render Webhook examples + frontmatter The page was already accurate (Webhook URL/Payload/Signing Secret/Headers, the automatic-headers table, HMAC details, outputs, POST-only callout, FAQ). Added frontmatter and rendered the two use cases as WorkflowPreview examples (notify a service, fire on a check). Adds WEBHOOK_NOTIFY_WORKFLOW and WEBHOOK_TRIGGER_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): add example + pageType to Workflow block reference The page was already accurate and well-structured (Configure It, outputs, deployment-status badge, execution notes, FAQ). Added pageType: reference and a rendered WorkflowPreview example showing a parent calling the child workflow enrich-lead and reading its result. Adds WORKFLOW_CALL_WORKFLOW. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): container rendering for Loop/Parallel + render the Loop example Adds subflow/container support to WorkflowPreview, modeled on the app's subflow-node.tsx: a solid-bordered box with a header (icon + name), an internal "Start" pill whose handle feeds the first nested block, and target/source handles at the vertical center. PreviewBlock gains size/parentId; edges gain an optional sourceHandle; nodes render nested children via React Flow parentNode. Renders the Loop reference's ForEach example (LOOP_WORKFLOW) and keeps the four loop-type sections + inside/outside referencing + caps. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): fix the Loop container's Start-pill connector The Start pill -> first-block edge wasn't rendering: it was a React Flow parent->child edge (unreliable), and the opaque container body hid it. Nested blocks now render as absolute-positioned top-level nodes (container below at zIndex 0, blocks above at zIndex 1), so the connector is an ordinary edge, and the container body is see-through so it's visible. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): render the Parallel example + frontmatter (last core block) Reuses the container rendering for the Parallel reference. Kept all substance (count/collection types, inside/outside referencing, batch size of 20, instance isolation, the Parallel-vs-Loop table, Best Practices, FAQ). Added frontmatter and a rendered container WorkflowPreview (PARALLEL_WORKFLOW: distribute tasks, call concurrently, aggregate <parallel.results>); the two use cases stay as labeled flows. Adds PARALLEL_WORKFLOW. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): prose glow-up for Guardrails to match the agent/condition voice Rewrote the listy register (**Use Cases:** / **How It Works:** / **Configuration:** scaffolding, "Use this when you need to..." filler) into the plain builder voice, matching the depth of the Agent/Condition/Function rewrites. Kept every validation type, option, range, the full PII entity/region list, the screenshot and video, the outputs table, the rendered examples, Best Practices, and FAQ. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): prose glow-up for Loop to match the agent/condition voice Rewrote into the plain builder voice and cut the filler: dropped the "Use this when you need to..." lines and the ASCII "Example: Iteration 1, 2, 3" pseudo-code, and folded the duplicated Inputs/Outputs tabs into Configuration + Referencing sections. Kept all four loop types with their screenshots, the inside/outside reference rules, the 1,000-iteration cap, sequential-vs-parallel guidance, the rendered example, and FAQ. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): prose glow-up for Parallel to match the agent/condition voice Same treatment as Loop: plain builder voice, dropped the ASCII pseudo-code and the duplicated Inputs/Outputs tabs, folded the verbose Advanced Features into tight Configuration + Referencing sections. Kept both types with screenshots, the batch-size-of-20 cap, instance isolation, large-result indexing, the Parallel-vs-Loop table, the rendered example, and FAQ. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): prose glow-up for Human-in-the-Loop Tightened the register: folded the pause sentence into the intro, made the section headers consistent (Configuration, Outputs), converted the bold-list Block Outputs into a table, condensed the Notification channel bullets to a line, and renamed the second "Example" so it no longer collides with the rendered Examples. Kept all the substance — Display Data / Notification / Resume Form, the Approval Methods and API Execute Behavior tabs, the portal video, and FAQ. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): re-enrich Loop prose (fuller, explanatory — not terse) The first glow-up overcorrected into terse fragments. Restored proper docs-quality prose at the Agent/Condition level: each loop type now explains what it does, when to use it, and the relevant reference; Configuration, Referencing, nesting, and Best Practices give context and the "why," not just bullets. Same substance, readable depth. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): re-balance Parallel prose to the Agent/Condition register Calibrated to the level signed off on elsewhere: each concept explained in a couple of clear sentences with a concrete detail — informative, not terse, not padded. Kept both types with screenshots, batch-size cap, isolation, large-result indexing, the Parallel-vs-Loop table, the rendered example, and FAQ. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): restore the Notification channel detail on HITL The glow-up over-compressed: it flattened the five notification channels (each with what they do) into one sentence. Restored them as a list in plain voice — tightening register shouldn't drop genuinely useful reference detail. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): builder-voice polish on the Credential intro Light touch only — the page was already well-structured and explanatory, so just led the intro with what the block does (and bolded the name) to match the other references. No content changed elsewhere. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(triggers): rewrite Start trigger in the builder voice Tightened the register, swapped the <code>&lt;&gt;</code> noise for backticks, added pageType + an outputs table, and kept all substance: Input Format types, chat-only outputs (input/conversationId/files), the editor/API/chat tabs, and best practices. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(triggers): rewrite Schedule trigger in the builder voice Plain voice and clean markdown (dropped the raw <ul>/<div> lists). Kept all substance: simple intervals, cron examples, timezone, deploy-tied activation, the 100-failure auto-disable, and FAQ. Added pageType. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(triggers): refocus Webhook trigger on the generic (native) trigger Rewrote in the builder voice and separated out the integration content: the page now documents the generic Webhook trigger (URL, Input Format, auth, custom response, outputs, dedup/rate-limit/deploy/no-auto-disable). The "trigger mode for service blocks" section is reduced to a short pointer + the demo video, and the long supported-services catalog and vague use-case bullets are dropped in favor of the Triggers index. Fixed the title (Webhook) and added pageType. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(triggers): builder-voice glow-up for RSS Light pass: added pageType + description, tightened the intro, and presented the output fields as an <rss.*> outputs table. Kept the polling config, use cases, the published-after-save callout, and the FAQ (poll cadence, dedup, 25-item cap, auto-disable, Atom support). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(triggers): rewrite Table trigger off the auto-generated card Replaced the BlockInfoCard/'provides 1 trigger' auto-gen format with a real builder-voice page: a spec-driven BlockPreview hero (added a 'table' spec), plain-language Configuration (table, event type, watch columns, include headers), and a full <table.*> outputs table. pageType: reference. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(triggers): frame the index around native triggers + separate the catalog Reframed "generic" as native (no connected account) and promoted RSS and Table into the native set alongside Start/Schedule/Webhook — cards, comparison table, and integration paragraph updated to match. In the sidebar, grouped the five native triggers under a "Native triggers" header and divided the ~44 service triggers under "Integration triggers" (nav-only — no files moved, URLs stable; the move to integrations/ is a later, separate change). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: promote Core Blocks + Core Triggers into the Workflows area Restructured the Documentation sidebar (meta-only — no files moved, URLs stable): after Deployment, the 16 core block pages now live under a "Core Blocks" section and the 5 native trigger pages under "Core Triggers", instead of buried in the bottom Reference catalog. Removed the now-redundant blocks tree from Reference, and retitled the Reference triggers tree "Integration triggers" so it holds just the service catalog (the native ones are promoted up top). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: merge block/trigger overviews into the Workflows overview; Core accordions Restructured the sidebar and overview hub (meta + content only, no integration files moved): - Folded the /blocks and /triggers overview pages into /workflows: the overview now carries the core-block catalog (do work / direct flow / shape run), the Integrations-and-triggers families framing, the native + integration trigger framing, the trigger comparison, manual-run priority, and email-polling groups. Deleted blocks/index.mdx and triggers/index.mdx as redundant. - Promoted the 16 core blocks into a "Core Blocks" folder accordion and the native triggers into a "Core Triggers" accordion, both under Workflows after Deployment. Integration triggers stay inside Core Triggers under a labeled divider, temporary until they move to integrations/<service> (tabs) later. - Repointed every /blocks and /triggers index link to the /workflows#blocks and /workflows#triggers sections. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: split integration triggers into their own Reference accordion Core Triggers is now the 5 native triggers only. Moved the 43 service triggers out of triggers/ into a new integration-triggers/ folder, surfaced as an "Integration triggers" accordion under Reference (an accordion must be its own folder in Fumadocs). In Workflows, Core Triggers now sits before Core Blocks. URLs: /triggers/<service> -> /integration-triggers/<service> (native /triggers/* unchanged); the integrations/<service> tabbed-page migration remains the later step. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): trim the overview back to an introduction It had drifted from a concept intro into a catalog. Kept the spine (the four parts with their previews, how-it-runs, workflows-in-context) and compressed the merged-in material: the full 16-block enumeration becomes a three-kind taxonomy with examples, the trigger section a short native/integration framing. Cut the anxious in-between — manual-run trigger priority, the niche email-polling-groups feature (belongs on the Gmail/Outlook trigger pages), the redundant block-def line, the Start-outputs callout half, the connections video, and the catalog-y FAQ items. Dropped the unused Video import. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: relocate email-polling + trigger-priority out of the overview Moved the two bits cut from the workflows overview to durable, generator-safe homes: email-polling groups -> the Integrations (connecting accounts) page; manual-run trigger priority -> the Start trigger page. Also added 'table' to the generator's HANDWRITTEN_TRIGGER_DOCS / SKIP_TRIGGER_PROVIDERS so the hand-written Table trigger page is no longer overwritten by generate-docs. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * feat(docs-gen): emit per-service integration pages (actions + Trigger section) Rewrites the generator to output one page per service under integrations/ instead of split tools/ + triggers/. Block pass writes the service's actions; trigger pass appends a '## Triggers' section (badged) to the same page, or writes a standalone page for trigger-only services. Meta is written after both passes; hand-written integration pages are preserved; docsUrl repointed to /integrations. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * feat(docs): unify tools + triggers into per-service /integrations pages Encodes the ontology "everything is a block; some blocks are triggers." The generator now emits one page per service under integrations/ — the service's Actions plus, when it has one, a Triggers section on the same page — replacing the split tools/<service> + triggers/<service>. No "Tools" terminology. - generate-docs.ts: output to integrations/, merge trigger sections into each service page (standalone for trigger-only services), Actions heading, table block now generated, docsUrl -> /integrations, hand-written pages preserved. - Nuked tools/ (213) and the interim integration-triggers/ (43); moved the custom-tools guide to building-agents/; knowledge/memory/file/table links and meta repointed to /integrations. - Sidebar: integrations catalog now under Reference (was tools); removed the Workspaces integrations entry and the integration-triggers tree. - block-icons: wait uses lucide Clock (the generated icons.tsx no longer carries a hand-added WaitIcon). Landing integrations data regenerated. No redirects (fresh start). Native Core Blocks/Core Triggers unchanged. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs): recover the hand-written manual-content intros on integration pages The tools->integrations relocation generated fresh pages, so the generator never saw the old tools/<service>.mdx to preserve its {/* MANUAL-CONTENT */} sections — 198 curated intros (AgentMail, etc.) were dropped. Reseeded each integrations page from the pre-move tools page in git, re-ran the generator (which now merges the manual intro into the new Actions/Triggers format), and repointed /tools/ links inside the recovered prose. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(scripts): rewrite the generator README for the integrations model Brings scripts/README.md current: integration pages are derived from the apps/sim block/tool/trigger registry (canonical-sources map), the golden rule not to hand-edit generated pages, the MANUAL-CONTENT escape hatch, which pages are hand-written/skipped, and the icons.tsx-overwrite gotcha. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: regenerate integration docs from staging-synced apps/sim After merging staging, regenerated so the integration pages reflect current source: correct block colors/configs (e.g. Gmail #FFFFFF), the new integrations (sendblue, millionverifier, neverbounce, zerobounce), and staging's icon set. Pages for integrations staging hid are removed; manual-content intros preserved. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs-gen): don't let stale-doc cleanup delete hand-written integration pages Staging's cleanupStaleToolDocs removes any integrations/*.mdx that isn't a visible tools block — it only guarded `index`, so it deleted the hand-written google/atlassian service-account pages. Now guards all HANDWRITTEN_INTEGRATION_DOCS. Restored the two pages, and repointed /integrations/file links to /files (staging hides the file block, so it has no integration page). Note: staging recategorized a2a/mysql/postgresql tools -> 'blocks' (and hid file), so they correctly drop out of the integration catalog and are currently undocumented — an IA decision to revisit. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs-gen): stop cleanup/writer filter mismatch from eating manual content Comprehensive-review findings, all generator-consistency bugs: - cleanup used staging's isIntegrationBlock while the writer kept the legacy filter, so integrations/{knowledge,memory,table}.mdx were deleted then regenerated without their manual intros every run. Both now honor a shared NATIVE_RESOURCE_BLOCK_TYPES set; intros reseeded. - Trigger-only services (imap, circleback; category 'triggers') were likewise deleted each run; the canonical set now includes visible trigger-category blocks, the standalone writer preserves manual content, and their intros are reseeded. - Mapped jsm -> jira_service_management, so JSM triggers merge into the JSM integration page instead of an orphan jsm.mdx (removed). - Repointed lingering bare /tools links to /integrations; added missing pageType to integrations/index and building-agents/custom-tools. Double-regen is now churn-free (idempotent) with all manual content intact. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs): recover staging's enriched Table doc + never drop manual content The merge resolution deleted staging's relocated blocks/table.mdx, which carried substantial enrichment our integrations/table.mdx (reseeded from the older tools/ version) lacked: Creating Tables (column types/constraints), Filter Operators, Combining Filters, Sort Specification, Built-in Columns, Limits, and Notes. Recomposed integrations/table.mdx with that content — Creating Tables inside the intro manual section, the reference tail in a notes manual section. Generator fix uncovered en route: a manual section whose insertion anchor is missing in the generated markdown (e.g. notes with no "## Notes" heading) was silently dropped on regen. Unplaceable sections now append at the end instead — manual content is never lost. Verified idempotent across double regeneration. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workspaces): de-philosophize the fundamentals prose Rewrote in the plain register of the workflows overview: 'draws the boundary for access' / 'Nothing crosses the boundary' / 'follow the same edge' become direct statements (only members can access it; a workflow in one workspace cannot read a table in another). '## The boundary' is now '## Access and isolation'. All substance kept: every resource type, permission levels, personal/organization/grandfathered kinds, deployments callout, VISUAL markers. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * fix(docs): restore #blocks and #triggers anchors on the workflows overview The editorial trim renamed '## Blocks' -> '## Kinds of blocks' and '## Triggers' -> '## How a workflow starts', silently breaking the ten /workflows#blocks and /workflows#triggers anchor links pointed there when the old index pages were folded in. Pinned the original ids with explicit heading anchors. Found by the comparative prose review. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: restore the genuinely useful reference bits the rewrite dropped From the comparative prose review, restored in guidance register (no spec dumps): temperature tiers on Agent (low/middle/high with ranges), loop/parallel iteration references in the variables syntax-at-a-glance table, and a short "Test it" section on the Webhook trigger (curl + check the run in Logs). The fourth flagged loss (tag-resolver mechanics on connections) turned out to be already covered — name normalization, case-sensitive paths, missing-output behavior, and value formatting are all on the page; only the internal resolver precedence chain was dropped, deliberately. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): rework the Agent intro — encyclopedia register Replaced the flat opening with a denser, factual one (no metaphor): what the block does, and its centrality stated as fact — 'Most workflows are built around one or more Agent blocks.' The agent-vs-Agent-block disambiguation moves from an info callout into a second paragraph on the block's role in building agents. Dropped the now-unused Callout import. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(integrations): add the HubSpot setup guide for the Marketplace listing Addresses HubSpot Marketplace review item A1: a public, HubSpot-specific setup guide following their template — what the app does, install + connect through the current flow (sidebar Integrations page -> HubSpot -> Add to Sim -> connect dialog -> HubSpot OAuth), with real screenshots of each step and a placeholder for the scope-approval shot; configure in a workflow (one-click skills/templates + the HubSpot block + trigger mode), use, disconnect (with data consequences), uninstall from the HubSpot side, troubleshooting. Capability wording is by CRM object rather than scope enumeration, so it stays accurate after the A2 scope trim. Lives at /integrations/hubspot-setup, guarded as hand-written, cross-linked from the HubSpot reference page's intro. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(integrations): rewrite the Integrations guide for the sidebar flow Integrations moved out of Settings to a top-level sidebar page. Rewrote the guide to the current journey: the Integrations page (Connected/Featured/search), service pages with one-click skills and templates, + Add to Sim -> connect dialog (display name + permissions) -> provider OAuth. Replaced the four Settings-era screenshots with current captures (connect dialog illustrated via HubSpot); block-side screenshots (account selector, manual credential ID) kept; one VISUAL marker for the connection detail view pending a fresh capture. Members/roles, credential-ID, reconnect/disconnect, email polling, and FAQ substance unchanged apart from navigation. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: move Building agents directly after Workflows in the sidebar The agent-building journey follows straight from workflows (blocks, triggers, deployment) rather than after the tour of every resource type. Tables/Files/ Knowledge Bases/Logs now follow it. Meta-only reorder. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: fill the visual slots coverable by existing components Six VISUAL markers filled with no new captures needed: - building-agents overview: rendered the minimal lead-scoring agent (Start -> Agent with tool chips -> Response, Agent highlighted) as a WorkflowPreview (BUILD_AGENT_WORKFLOW) - files guide: the read -> summarize -> write chain as a WorkflowPreview (FILE_SUMMARY_WORKFLOW) - tables guide: the query -> classify -> write-back roundtrip as a WorkflowPreview (TABLE_ROUNDTRIP_WORKFLOW) - choosing guide: the six-kind comparison grid as a markdown table - knowledgebase guide: the Knowledge block's output as an OutputBundle - workspace fundamentals: removed a duplicate nesting-diagram marker 42 -> 39 VISUAL markers remaining (screenshots + designed diagrams). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(components): run-inspector OutputBundle + lightbox with block inspector Two visual-component upgrades, both mirroring the real app: - OutputBundle is now a miniature of the run inspector: a Logs column (block rows with icon chips and durations, source selected) beside the Output panel's typed tree — keys with the app's type-badge semantics (string green, number blue, object gray, array purple, boolean orange), chevrons, indent guides, primitive values. Styling lifted from the terminal's structured-output. Dropped the "Read one value by name" footer (the prose teaches the tag). The three usages (data-flow, tables, knowledgebase) get real typed trees; data-flow's stale purple/gpt-4o example corrected en route. - WorkflowPreview gains a lightbox + read-only block inspector: clicking a block (or the expand control) opens a 92vw/86vh overlay with zoom and pan, and a right-hand inspector panel showing the selected block's full configuration — canvas rows truncate, the inspector doesn't. Fields render as app-style controls (dropdown/textarea/input by heuristic) with dashed dividers, tool chips, and a Connections footer computed from the edges. Selection rings without dimming (new selectedBlock option in workflow-data). Esc/backdrop closes; body scroll locks while open. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: regenerate after staging merge — AppConfig joins integrations/ Staging's new AWS AppConfig integration (#4928) generated its docs into the old tools/ layout; re-homed to integrations/appconfig.mdx (Actions heading, meta entry) via the generator. tools/ stays deleted. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: redirect the retired tools/ and trigger URLs to integrations/ Revises the earlier fresh-start call: /tools/* are ~200 live, indexed URLs referenced by deployed app versions' docsLink fields and marketplace listings, so dropping them cold would 404 from the live product. next.config now 308s: - /tools -> /integrations, /tools/:slug -> /integrations/:slug (custom-tools -> building-agents/custom-tools first) - old /triggers/<service> -> /integrations/<service>, enumerated so the native trigger pages keep resolving; provider-slug mappings for jsm and the hyphenated Google/Microsoft slugs - /blocks and /triggers index URLs -> the workflows overview anchors Verified every class + native passthroughs against the dev server. Spec updated. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(getting-started): rewrite — current UI, cut the post-tutorial padding The last old-guard page. Accuracy: Agent config now uses Messages (System/User message) instead of the removed System Prompt/User Prompt fields, the default model instead of GPT-4o, the banned 'no-code' phrasing is gone, the deploy card points at /deployment, and frontmatter gets description + pageType. Weight: cut the 'What You've Built' checklist, the 'Key Concepts You Learned' re-teach section, the duplicate 'Resources' links, the Start-block hand-holding, and ten dead icon imports; tightened every step preamble. 203 -> 113 lines with the full 5-step tutorial, videos, and FAQ intact. (Videos still show the old UI until the re-recording pass.) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: de-fluff the Tier-1 heavy pages (logging, mcp, passing-files, permissions) From the exhaustive fluff audit, keeping all substance: - logging: merged the duplicated Console/Logs-page structure, snapshot concept stated once instead of three times, cut the generic Best Practices, trivial tab walkthrough condensed. Frontmatter added. - mcp: intro + "What is MCP?" generic bullets folded into two sentences, cut the Common Use Cases catalog and the verify-your-config Troubleshooting checklists, merged the twice-stated Refresh behavior, security kept as one real warning. - passing-files: marketing opener replaced with a factual lead, fixed the stale retired-Vision-block reference (now Agent with a vision model), dropped the FAQ item that restated the block catalog verbatim. - permissions: heading-restating intro replaced with the two-layer model, cut the three "Perfect for: stakeholders..." persona lines and the generic Best Practices section, dropped the FAQ restating the limits table. - connectors: audit over-flagged it — the categorized support matrix, API-key table, and config examples are genuine reference; frontmatter only. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: tier-2 fluff trims (costs, enterprise, mailer, skills) Conservative sweep from the audit, unambiguous cuts only: the costs CYA opener and formula restatement, the enterprise marketing intro (now a functional summary), mailer's restated convenience line and chat-upload comparison, and skills' third restatement of progressive disclosure. Audit flags screened out as misfires: mothership/tasks (immediate-vs-scheduled are two facts, not a duplicate), self-hosting telemetry (real sizing data), and the recently approved credential/HITL/workflow-block/trigger pages. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(skills): update to the Skills tab on the Integrations page + document import Skills moved again — they now live on the Integrations page's Skills tab in the workspace sidebar (the doc said "Open the Skills page"). Updated the create flow (+ Add to Sim -> Add Skill dialog) with fresh screenshots of the tab and both dialog tabs, and documented the previously-missing Import flow: upload a .md with YAML frontmatter or a .zip containing SKILL.md, fetch from a GitHub URL, or paste SKILL.md content (verified against the import route/component; name 64 / description 1024 limits verified against the contract). Noted the curated skills suggested on integration pages, cross-linked the Skills tab from the Integrations guide, and refreshed the location FAQ. Mechanics (progressive disclosure, load_skill, agent-block attachment) unchanged and still accurate. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(building-agents): render the lead-scorer running example on choosing The page narrated its running example through six sections without ever showing it. Authored LEAD_SCORER_WORKFLOW (Start -> Enrich workflow-as-tool -> Function reshape -> Agent with Search/Send Email/CRM tool chips -> Google Sheets append) and rendered it after the intro, with highlightBlock re-renders in the three sections that map to a node (deterministic block -> the Sheets append, agent tool -> the Agent, workflow-as-tool -> Enrich) — the same pattern as the workflows overview. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(tables): rewrite workflow columns around the real lead-scoring example Rebuilt the page on the ai_startup_customers screenshots instead of captioning them onto the old hypothetical: one running example throughout — Company Domain fills domain, Company Info reads it into employee_count/description, Lead Score Enrichment writes lead_score/priority/score_reasoning. Every section now describes the actual UI: the grid with group headers, per-row run buttons, and the 21-running toolbar; the Configure workflow panel (picker, column inputs, output selection, Auto-run, Run after); the Company Info input/output mapping; Not found cells explained where the screenshot shows them; the cascade section describes the example itself. All placeholder markers on the page resolved. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: regenerate after staging merge — Slack trigger update + file block re-visible Staging's mothership v0.2 (#4923) expanded the Slack trigger payload (interactivity, slash commands: event_type, command, action_id/value/actions, response_url, trigger_id, callback_id, ...) — regenerated so it lands on the unified integrations/slack page; the old-layout triggers/slack.mdx from staging's generator was dropped in the merge. The file block is visible again upstream, so integrations/file.mdx is back in the catalog. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(tables): playbook prose pass on workflow columns + restore File block links Workflow columns, against the docs-writing playbook: killed the banned 'Term — desc' bullets in the Configure list (term + verb form), restored the one universal analog (spreadsheet macro), fixed the clipped 'On,/Off,' fragments, replaced an invented <start.companyDomain> tag with the verified description, and thinned em-dashes to four page-wide with no clustering. Also repointed [File] block mentions back to /integrations/file now that the page exists again (FileV5 is visible upstream); the Files-store links stay on /files. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(tables): per-row execution inspection on workflow columns Two new captures: the cell menu (View execution, Re-run cell, row actions) and the Log Details trace for a single row's run. New 'Inspecting a row's run' section ties cell values to real, traceable runs; corrected the re-run guidance now that Re-run cell exists (the page previously said Run all rows was the only way to retry). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): drop the confusing 'order by hand' sentence 'You never set the order by hand' read wrong (wiring connections is setting it by hand), and the replacement was over-explanation. The first sentence already carries it: Sim works out the order from the connections. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): fix the over-claim about independent blocks 'Two blocks that don't depend on each other run at the same time' is wrong — independent blocks at different depths run at different times. Concurrency follows from readiness, not independence: blocks whose dependencies have all finished run together. Reworded to say that, tied to the image's two agents. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): accuracy audit of how-it-runs against the executor Verified every claim on the page against apps/sim/executor. One claim was materially false: "a failed block stops its own path but leaves independent paths running" — in the engine, an unhandled block failure sets the error flag and stops scheduling entirely (in-flight blocks finish, nothing new starts); only a connected error port routes the failure and keeps the run alive. Now says that. Two imprecisions tightened: a join waits for every feeder *that is going to run* (deactivated-branch feeders don't hold it up, per the edge-manager cascade), and Loop also repeats while a condition holds. Confirmed accurate: per-block readiness scheduling (readyQueue + race, not layers), branch-skip cascade and empty tags, the 25-hop call-chain cap. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(logs): real captures on the overview + prose matched to the UI The logs-debugging overview had six visual placeholders and no visuals. Three real captures placed: the workspace Logs page as the hero (rows with status, credits, trigger, duration), Log Details' Trace tab at the blocks section (the CRM sync run's spans, with a one-line read of where the time went), and the editor's live run console at the input/output section. Prose corrected to what the UI shows: cost is in credits, failed runs are badged Error (dropped the five-state enum the list doesn't display), and the Trace tab is named. The row-anatomy marker is covered by the hero; the two designed-diagram markers (debug-loop flowchart, failed-vs-success comparison) remain. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): one reference syntax, named sources — untangle variables vs connection tags An exhaustive sweep of "connection tag" found the docs asserting both that a workflow variable is a connection tag (response.mdx used it as the umbrella for all angle-bracket references) and that it isn't (variables.mdx). Ruled the narrow definition canonical — a connection tag reads a block's output; the name follows the connection — and restructured around the real model: - variables.mdx: new "One syntax, named sources" section states that everything in angle brackets is one mechanism whose first segment names the source, with the load-bearing fact stated plainly: `variable` is literal, a connection tag starts with the block's own name. The syntax table drops the redundant dot-notation row, gets one row per source, and is ordered by resolution precedence with the order explained beneath it (absorbing the old Name conflicts section). The credentials pointer folds into the env-var section; trimmed the "never appears in outputs" overclaim. - response.mdx: no longer calls a workflow variable a connection tag. - connections.mdx: the owner page closes the loop — same syntax also reads variables and loop/parallel context; a connection tag is the block-output case. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): verify the reference model against the resolver; fix one imprecision Checked every claim in the new 'One syntax, named sources' section against apps/sim/executor/variables: resolver chain order is Loop -> Parallel -> WorkflowVariables -> Env -> Block (matches the table); 'variable'/'loop'/ 'parallel' are literal prefixes (REFERENCE.PREFIX); block names normalize via toLowerCase + strip spaces; an unmatched reference is genuinely left in place (resolver returns undefined -> the replacer emits the raw match). One claim tightened: {{KEY}} is a different syntax and can never collide with angle-bracket references, so the precedence sentence now scopes collisions to the angle-bracket sources with a concrete example (a block named 'variable'). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(building-agents): workflow-as-tool is agent-decided, not the Workflow block The choosing page defined workflow-as-tool as the Workflow block (path-decided), contradicting its own name and the comparison table's premise. Verified against the product: workflow_executor is an agent tool — you pick the workflow in the Agent block's tool list, the model decides when to call it and supplies the inputMapping (user-or-llm), inputs arrive at the child's Start trigger. Rewritten agent-first: the section defines it as a workflow handed to an agent as one callable tool, the lead scorer gains a Deep Enrich workflow tool chip on the agent (diagram updated), and the deterministic Workflow block becomes the explicit contrast in a callout — same child workflow, the difference is who decides, mirroring the block/agent-tool contrast. Table row corrected to "The agent"; the summary paragraph follows. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: theme-aware previews + enrichments vs workflow groups split Light-mode support for every preview component (WorkflowPreview canvas, nodes, containers, edges, lightbox, BlockPreview, OutputBundle, BlockInspector): a wp-scope token block in the docs global stylesheet whose values mirror the OG repository's globals.css in both modes (surfaces, borders, --workflow-edge, text tiers, the --badge-* type-badge palette). Every hardcoded hex swapped to a --wp-* var; brand colors, selection blue, and error red stay literal. tables/workflow-columns: separated the two group kinds per the contract's workflowGroupType enum ('manual' | 'enrichment'). New "Two kinds of groups" section opens with the + New column menu capture (Enrichments above the types, Workflow below); Enrichments documented from the code-defined registry (company domain, company info, email verification, phone number, work email) including the provider-cascade behavior that produces Not found cells; the Company Info panel capture is now correctly labeled as an enrichment config; workflow groups keep the Configure workflow panel. Shared machinery generalized under "How groups run"; the cascade section names which stage is which kind; the two portrait screenshots render smaller. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(tables): don't enumerate the enrichment catalog; don't assert a group's kind Two corrections on workflow columns: the prose no longer lists the enrichment catalog (growable, not procedurally tracked — it now describes the category and points at the Enrichments panel; the provider-cascade/Not-found explanation stays, it's behavior not catalog), and the page no longer asserts which kind the example's Company Domain / Company Info groups are (Company Info may be a user-built workflow, not the built-in). The input/output bindings capture moved to "How groups run" as the kind-agnostic illustration; only Lead Score — whose panel shows the workflow picker — is named as a workflow group. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(components): per-branch source handles — conditions and routers finally branch WorkflowPreview's node only ever had one header source handle, so every condition/router example fanned both edges out of a single point and never showed the if/else rows the real canvas (and the BlockPreview hero specs) render. PreviewBlock now supports `branches` (each rendered as a row with its own right-edge source handle, id `branch-<id>`) and `showError` (red error handle), mirroring the executor's per-branch condition-true/condition-false and router-<route> handle model. A block with branches emits from them, not the header. Every affected example rewired (13 workflows): the three condition examples, status-per-branch, credential routing, and the webhook-trigger check route their edges through branch-if/branch-else with the expression on the If row and an explicit else; the three router examples list their actual routes as branch rows (Sales/Support/Billing, Product/Bug report, Enterprise/Self-serve); the terminal gates (variables retry, evaluator gate, the three guardrails gates) show dangling if/else branch rows like the canvas does. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(components): inspector shows branch rows Moving condition expressions from rows into branches emptied the lightbox inspector for condition/router blocks — it only mapped rows to fields. Branches now map too: each branch renders as a field (If with its expression as code, else as an empty control, router routes by name). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(components): branch handle ids match the app's workflow representation Verified against the source after the branch-handles work: the canvas emits condition-${cond.id} handles per condition row (workflow-block.tsx) and Router V2 uses router-${routeId} port handles, and edges carry those ids as sourceHandle — the docs' invented branch- prefix was a gratuitous divergence that the planned fromWorkflowState() adapter would have had to translate. The node now uses the authored branch id as the handle id directly, and every example authors ids in the app's own scheme (condition-if/condition-else, router-<route>), so example edges now match real workflow edges verbatim. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: agent skills mint /integrations/ docs links and describe the new output The add-integration/add-block/validate-integration skills — what Claude Code follows when integrations land on staging — still taught the old layout: docsLink templates pointing at docs.sim.ai/tools/{service} and 'generates tools/{service}.mdx'. Updated so that once this PR merges, the instructions on staging produce the new way by themselves: /integrations/ docsLinks, the per-service page description, and the don't-hand-edit/manual-content pointer. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): execution semantics, not simultaneity The concurrency section drifted into 'run at the same time' framing across two accuracy passes — but the semantics are non-blocking execution: a block starts the moment its dependencies finish and waits on nothing else. Section retitled 'Blocks run as soon as they can', the rule stated in two plain sentences, the duplicated pre-image example narration gone (the post-image caption carries it). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): errors are execution semantics — own section on how-it-runs Failure behavior was buried inside 'Watching a run' (the live-UI section). Now a first-class 'When a block fails' section in the execution story: an error fails the run (in-flight blocks finish, nothing new starts) unless the block's error port is connected, in which case the run follows the error path. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: data-driven additions from the platform-metrics read Three targeted edits from the sim-internals analysis, each carrying an inline {/* why */} provenance comment so future editorial passes know the data behind it: - workflows/how-it-runs gains "How long a run can take" — run timeouts are the only hard-error class provable at scale (2,415 five-minute timeouts in 14 days); limits verified in lib/core/execution-limits/types.ts (5 min free / 50 min paid sync, 90 min async, env-overridable). - getting-started gains an "if the run doesn't go green" callout at the Test step — the largest funnel drop is created-workflow -> first-successful-run (92% -> 49%), and this is the stall point. - function/api Best Practices: the existing error-path bullets get a guard comment (<1% of deployed workflows connect an error port — under-adopted, not under-needed) instead of duplicate bullets. - visuals manifest: capture priority reordered by integration adoption (Sheets, Gmail, Telegram, WhatsApp, ...). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: regenerate after staging merge (integration validation batch + Gong tools) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: rename Building agents -> Agents; URLs match the settled IA The section's pages now live where the sidebar says they do: building-agents/ -> agents/, and the stray top-level /mcp and /skills fold in as /agents/mcp and /agents/skills (they were always part of the agents story — the URLs predated the IA settling). Sidebar section header is now "Agents", link labels updated, and every old URL 308s: /building-agents(/*) -> /agents(/*), /mcp, /skills, plus the existing capabilities/ and tools/custom-tools redirect destinations retargeted. Verified: all five new pages render and every old-URL class redirects correctly. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(workflows): connections gets its video, an FAQ, and accurate output examples The reorg dropped two things from the old tags page that belonged on the connections reference: the connections.mp4 walkthrough (restored after the intro) and the FAQ (rebuilt in the robust JSX form — resolver order, name normalization, env-var syntax pointer, didn't-run behavior, array indexing, Function-block formatting; answers aligned with the since-verified resolver facts, including unmatched-references-left-in-place). Editorial/accuracy pass on the output-shape tabs while in there: stale gpt-4o and gpt-5 examples now claude-sonnet-4-6, the Agent tokens shape corrected to the verified { input, output, total } (the page contradicted blocks/agent), and the dubious cost: [] line dropped — the example now matches the real run inspector. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: regenerate after staging merge — sim trigger, enrichment + logs blocks, re-shown DB integrations Staging's #4941 added the Sim workspace-event trigger (hand-written page adopted into Core Triggers), the Enrichment and Logs blocks (category 'blocks' — added to NATIVE_RESOURCE_BLOCK_TYPES so they live in the integrations catalog like table/knowledge/memory), and re-categorized mysql/postgresql/sftp/smtp/ssh back to visible tools (their pages return to the catalog). Generator sets merged as the union of both sides (sim in HANDWRITTEN_TRIGGER_DOCS + SKIP_TRIGGER_PROVIDERS, enrichment in the icon allowlist). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: regenerate after staging merge (CodePipeline); suppress sim trigger from catalog The native Sim workspace-event trigger is documented at triggers/sim — the block writer no longer emits an integrations page for it (skip + canonical-set exclusion). CodePipeline (#4945) lands in the catalog in the Actions format. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(blocks): cross-link the Memory block from the Agent memory section Final loss audit found the old page's pointer from built-in agent memory to the standalone Memory block had been dropped; one line restores it. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: URLs now mirror the sidebar — sections own their pages Every page lives at a path matching its meta.json section, done now while none of these URLs are publicly live (the last free window before merge): - Workflows owns its accordions: /blocks/* -> /workflows/blocks/*, /triggers/{start,schedule,webhook,rss,table,sim} -> /workflows/triggers/*, /deployment/* -> /workflows/deployment/* - Mothership owns Mailer: /mailer -> /mothership/mailer - Workspaces & Access folds into Platform, sequenced concept-first with the reference tail last: /platform/{workspaces,organization,permissions, credentials,costs}, then platform/self-hosting/*, platform/enterprise/* (from /workspaces/fundamentals+organization, /permissions/roles-and- permissions, /credentials, /costs, /self-hosting/*, /enterprise/*) All internal links swept (0 broken in a full-tree resolver sweep), root meta.json repointed, and every previously-live URL 308s to its new home — including retargeted destinations of existing redirects so chains stay single-hop (verified: /execution/chat reaches /workflows/deployment/chat in one hop), and the native-trigger rule ordered after the enumerated integration-trigger redirects so /triggers/gmail still reaches /integrations/gmail. Production build passes. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: untrack .plans/ (local agent planning files) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(preview): tool chips use the EMCN ChipTag chrome The canvas previews' tool chips were ad-hoc (5px radius, header surface, plain border). The app's canonical chip chrome is the ChipTag family: 20px tall, rounded-md, px-1, gap-1.5, --surface-5 light / --surface-4 dark with an inset --border-1 ring and --text-body label. Mirrored those values into --wp-chip-* tokens (both modes) and restyled the chip; the integration's brand-color icon square stays, sized to the chip. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * chore(blocks): one-time shift of all docsLinks to the new docs URLs Block definitions are the patterns coding agents copy from, so redirects alone leave new blocks minting dead conventions. Every docs.sim.ai link in apps/sim now points at the final URL scheme: /tools/<slug> -> /integrations/<slug> (433 links), /blocks/<core> -> /workflows/blocks/<core> (knowledge/enrichment/ logs -> /integrations/*), native /triggers/* -> /workflows/triggers/*, /mcp -> /agents/mcp, /self-hosting + /enterprise -> /platform/*, plus the llms.txt listings and the blocks.test.ts assertions. Verified every rewritten target against the docs tree: all resolve except ten hidden blocks (vision, spotify, thinking, tts...) and a2a whose links were already dead pre-reorg — no regressions introduced. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: ignore .plans/ (local agent planning files) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs(files): align every File-block claim with the shipped file_v5 block Accuracy audit against apps/sim/blocks/blocks/file.ts (FileV5Block, the visible block) and tools/file/*: - The block has FIVE operations, not four — Get Content was missing entirely. - Read outputs file objects only; the page claimed it also returned extracted text. Text comes from Get Content (contents, per file) or Fetch (combinedContent) — table, prose, and the Fetch callout corrected. - Functions CAN read files: sim.files.readText/readBase64 exist in the sandbox (isolated-vm-worker.cjs), so "doesn't reach into workspace storage" is gone; the section now teaches Get Content text or sim.files on the file object. - Workspace file IDs are wf_<shortId> (workspace-file-manager.ts:511), not f_. - Stale "such as Claude or GPT-4o" vision parenthetical dropped. - "File block reference" card pointed at /files (the section overview); now /integrations/file. - FILE_SUMMARY example agent consumed <file.combinedContent>, which Read never produces — now binds the file object to the Files input. - passing-files.mdx: combinedContent scoped to Fetch, contents documented. Verified intact: Write's numeric-suffix collision behavior, Fetch's auth headers, Append-by-name, and the file-object shape. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: keyboard-shortcuts audited against the command registry; cut legacy workspace detail Every binding verified against commands-utils.ts (the global registry), workflow.tsx, and table-grid.tsx. Three fixes: tables Mod+A (select all rows) doesn't exist — the real bindings are Shift+Space (select row, was misworded as a toggle) and the undocumented Mod+Space (select column); the global Mod+Shift+A row conflated two commands — add-agent (Mod+Shift+A) and add-workflow (Mod+Shift+P) are separate. All 29 other documented shortcuts confirmed accurate, including tables clipboard (native copy/cut/paste events) and Mod+Y redo (tables only — correctly absent from the workflow editor section). Also drops the grandfathered_shared workspace paragraph — internal billing taxonomy, not something a reader can act on. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: apply Theodore's accuracy feedback - getting-started: workflow creation is the + button next to Workflows in the sidebar (no "New Workflow" button exists); Exa/Linkup no longer need user-supplied API keys on hosted Sim (apiKey is hideWhenHosted in the Exa block) — step and FAQ updated. - workflows overview: chat and API are entry points of the Start trigger, not separate triggers — the "swap in a chat/API trigger" sentence now matches triggers/start's own model. - variables: names cannot contain periods — the resolver reads everything after the first dot as a path into the value (executor/variables/resolvers/ workflow.ts splits on dots) — constraint now stated where name normalization is taught. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * docs: Python sandbox package list (verified) + agent/agents cross-linking Function block: the Python callout's "common packages like matplotlib" becomes the actual package list, grouped by use. Sources verified 2026-06-10 and cited in an inline provenance comment: E2B's code-interpreter template requirements (the base Sim's mothership-shell template builds from) plus Sim's three pip additions (awscli/yq/csvkit, per the copilot repo's template.ts via sim-internals). Versions omitted so the list doesn't rot on routine bumps. Agent surfaces deduplicated by direction: blocks/agent's Tools section now links custom tools and MCP and points at the Agents concept page for tool sourcing; agents/index drops its duplicated Auto/Force/None enumeration in favor of the block reference, which owns config mechanics. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	2026-06-10 18:39:57 -07:00
WaleedandTheodore Li	147ac89672	feat(docs): fill documentation gaps across platform features (#4110) ... * feat(docs): fill documentation gaps across platform features * fix(docs): address PR review comments on chat OTP cookies and MCP env var placeholders * fix(docs): replace smart quotes with straight quotes in JSX attributes * update(docs): update mcp, custom tools, and variables docs * Fix grammar * mothership docs, tags, connectors, api, chat deploy, etc * more info * more * feat(docs): auto-generate per-provider trigger documentation Extends scripts/generate-docs.ts to produce one MDX page per trigger provider (39 pages) in apps/docs/content/docs/en/triggers/. The 5 hand-written pages (index, start, schedule, webhook, rss) are never touched. Key additions to the generation script: - resolveConstVariable() resolves module-level const spreads so providers like Vercel that build outputs from const variables (not just functions) are fully documented - resolveTriggerBuilderFunction() extended to expand variable spreads (...varName) in addition to function-call spreads (...fn()) - groupTriggersByProvider() deduplicates v1/v2 trigger variants by name, keeping the highest-versioned one per provider - writeIconMapping() adds bare-name aliases for versioned block types (github_v2 → github, fireflies_v2 → fireflies, etc.) so BlockInfoCard resolves icons for all 39 trigger providers - extractTriggerConfigFields() filters readOnly display blocks (webhook URL displays, sample payloads, curl examples) from config tables Each generated page includes: BlockInfoCard with correct icon/color, trigger count, polling note where applicable, Configuration table, and Output table for every trigger. No "Type:" lines. * refactor(docs): align trigger docs structure with tools docs - Use ### `trigger_id` headings (matching ### `tool_id` in tools docs) - Wrap all trigger sections under a ## Triggers header - Rename Configuration/Output to #### level (matching #### Input/Output) - Use Parameter column header to match tools docs table style - Map UI widget types to semantic types: short-input/long-input/dropdown → string, switch → boolean, slider → number, oauth-input → string * refactor(docs): use human-readable names for trigger section headings Trigger IDs are internal identifiers; users scan by name. Switch from ### `trigger_id` to ### Trigger Name for cleaner sidebar navigation and better readability. * fix(docs): resolve subBlock builder functions for all trigger Config sections Extends generate-docs.ts to parse subBlock builder functions so all 15 providers previously missing Configuration sections now generate them. Handles three patterns: - `buildTriggerSubBlocks({extraFields: buildX(...)})` — extracts extra fields from the call site and resolves them from the provider's utils.ts - `return [...]` — direct array return (Attio, Confluence, etc.) - `blocks.push(...)` — imperative push pattern (Linear, Ashby) Also resolves const-reference field IDs (SCREAMING_CASE) by searching the webhook provider constants cache, fixing Gong's `gongJwtPublicKeyPem` field which was previously unresolvable. Adds title-as-description fallback for OAuth credential fields that have no explicit description. * fix(docs): correctly destructure nested implicit-object trigger outputs Fixes a parser bug where output fields with no top-level `type` key but child fields each having their own `type`/`description` were incorrectly parsed. The `type:` and `description:` regex matches were not depth-aware, so values from nested children bled into the parent field. Changes: - Add `isAtDepthZero()` helper for brace-depth-aware regex matching - Fix `parseFieldContent` to only match `type:` at brace depth 0 - Fix `extractDescription` to only match `description:` at brace depth 0 - Add implicit-object fallback: when no top-level `type` exists but child fields have their own types, treat as `object` with `properties` - Regenerate all affected trigger docs (Cal.com payload, Linear data, Jira issue.fields, Ashby application, Greenhouse candidate, etc.) * chore(docs): update static trigger and start page images * feat(providers): add claude-opus-4-7 model with adaptive thinking support * Add workflow version screenshots * Add function block screenshots --------- Co-authored-by: Theodore Li <theo@sim.ai>	2026-04-16 11:51:49 -07:00
Emir Karabeg	ad100fa871	improvement(docs): ui/ux cleanup (#4016) ... * improvement(landing, blog): SEO and GEO optimization * improvement(docs): ui/ux cleanup * chore(blog): remove unused buildBlogJsonLd export and wordCount schema field * fix(blog): stack related posts vertically on mobile and fill all suggestion slots - Add flex-col sm:flex-row and matching border classes to related posts nav for consistent mobile stacking with the main blog page - Remove score > 0 filter in getRelatedPosts so it falls back to recent posts when there aren't enough tag matches - Align description text color with main page cards	2026-04-07 11:05:58 -07:00
Waleed	dcbdcb43aa	chore(deps): upgrade to nextjs 16 (#2203) ... * chore(deps): upgrade to nextjs 16 * upgraded fumadocs * ensure vercel uses bun * fix build * fix bui;d * remove redundant vercel.json	2025-12-04 17:55:37 -08:00
Waleedandwaleedlatif1	5d791cd55f	feat(i18n): update translations (#2178) ... * feat(i18n): update translations * memory optimizations --------- Co-authored-by: waleedlatif1 <waleedlatif1@users.noreply.github.com>	2025-12-04 00:18:21 -08:00
Aditya Tripathi	aa2577be91	feat(llms): add LLM text processing and routing for MDX and TXT files (#470)	2025-06-10 08:32:37 -07:00
Waleed Latif	a92ee8bf46	feat(turbo): restructured repo to be a standard turborepo monorepo (#341) ... * added turborepo * finished turbo migration * updated gitignore * use dotenv & run format * fixed error in docs * remove standalone deployment in prod * fix ts error, remove ignore ts errors during build * added formatter to the end of the docs generator	2025-05-09 21:45:49 -07:00