docs(approval-engine,memory): align beta.31 with house style and add Memory overview#199
Merged
Merged
Conversation
Redundant with docs-sync-skills which handles syncing on push to main. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Quickstart for building bots on Velt comment threads with the Chat SDK and @velt-js/chat-sdk-adapter. Wired into the Integrations nav group. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…-bot Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…rse error) scripts/skills-sync/reference-artifacts/format-guide.md is a template artifact, not a docs page; its <optional ...**bold**...> syntax broke the MDX parser and failed the deploy. Excluding scripts/** from the Mintlify build. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The docs.json 'ignore' field is not honored by Mintlify for page parsing; the build explicitly fetches .mintignore. Excludes scripts/ (skills-sync pipeline + reference-artifacts/format-guide.md) whose template syntax broke the MDX parser. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Retry after 'Failed to revalidate subdomain: velt' on prior deploy. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
…lidation failure scripts/ files (incl. format-guide.md) predate the last known-good deploy (98c6ed0), so the .mintignore could not have fixed a real new break. Reverting to test whether the .mintignore itself triggers 'Failed to revalidate subdomain'. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Isolates the 'Failed to revalidate subdomain: velt' failure. Restores .mintignore (proven to fix the format-guide.md parse error) and removes the chat-sdk page so docs content matches the last known-good deploy byte-for-byte. If revalidation STILL fails here, the cause is Mintlify-side, not our commits. Chat-sdk page is preserved in history (3dbafaf) and will be re-added once the deploy is green. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…failure) Isolation test (99aa37d) proved the 'Failed to revalidate subdomain: velt' failure reproduces with docs content byte-identical to the last-good commit 98c6ed0 — so the page is not the cause. Restoring it and the AI-group nav entry. The revalidation failure is Mintlify-side and tracked separately. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Remove internal customPlanInfo fields from Create API Key docs
Co-authored-by: Cursor <cursoragent@cursor.com>
… docs Restructure setup and customize-behavior into method-named sections with React/Other Frameworks tabs, rename approved -> accepted lifecycle, and simplify the Suggestions changelog entry. Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com> # Conflicts: # api-reference/sdk/models/data-models.mdx
…h SDK Reviewed the Suggestions feature source in velt-js/sdk and sdk-react and corrected the docs to match shipped behavior: - customize-behavior.mdx: concept-first rewrite explaining the suggestion pipeline (enable -> snapshot -> diff -> pending -> accept/reject -> apply), getters and the edit-time-state gotcha, the three commit paths, applying accepted changes (with idempotency), and a lifecycle/events reference. - Fix registerTarget: it returns void (remove via unregisterTarget), not an unsubscribe function (setup.mdx, customize-behavior.mdx, api-methods.mdx). - Fix SuggestionStatus value accepted (was approved) and document apply_failed as implemented via ApprovedSuggestion (data-models.mdx). - Fix overview event name back to suggestionApproved (was suggestionAccepted). Co-authored-by: Cursor <cursoragent@cursor.com>
…ejected events
The accept/reject outcome is emitted on the comment element
(CommentEventTypes.SUGGESTION_ACCEPTED / SUGGESTION_REJECTED), not the
SuggestionElement: acceptSuggestion/rejectSuggestion flip annotation.type to
'comment' before SuggestionActionsService.projectAndTrigger, which early-returns
on non-suggestion type, so the SuggestionElement's suggestionApproved/
suggestionRejected never fire on resolve.
Switch setup, customize-behavior, and overview to the comment-element events
(useCommentEventCallback('suggestionAccepted') / commentElement.on(...)) and read
commentAnnotation.suggestion.newValue. SuggestionElement still owns create/read/
edit-capture plus suggestionCreated and suggestionStale.
Co-authored-by: Cursor <cursoragent@cursor.com>
Merge overview, setup, and customize-behavior into a single Suggestions page (overview.mdx): feature pitch, how-it-works pipeline, define targets, enable mode, capture options, apply-on-accept, read for custom UI, behavior notes (drift / stale-wins), lifecycle + events reference, and data model. - Delete setup.mdx and customize-behavior.mdx. - Collapse the docs.json Suggestions group to the single page. - Repoint Full Documentation links from /suggestions/setup to /suggestions/overview in api-methods.mdx (14) and react-hooks.mdx (11). Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Restructure the page so the how-to lives under one numbered Setup section (1. define targets, 2. enable mode, 3. capture edits, 4. apply on accept), folding the former standalone sections in as steps. Remove redundant behavior notes (one-per-session and mode-not-persisted are already covered inline) and move the get-SuggestionElement snippet into Setup. Co-authored-by: Cursor <cursoragent@cursor.com>
v5.0.2-beta.23
… references Port changes from raghul/agent-docs (PR #168) scoped to: - api-reference/rest-apis/v2/comments-feature - api-reference/rest-apis/v2/workspace Includes new activity/apikey/notification/permission-provider config endpoints and renames apikeymetadata-get to permissionproviderconfig-get. Co-authored-by: Cursor <cursoragent@cursor.com>
v5.0.2-beta.25
Co-authored-by: Cursor <cursoragent@cursor.com> # Conflicts: # release-notes/version-5/sdk-changelog.mdx
v5.0.2-beta.26
* v5.0.2-beta.26 * v5.0.2-beta.27 * updaet --------- Co-authored-by: Rakesh <rakesh@snippyly.com>
…skills to dispatcher, bump Agent 2 turns - Add api-reference/sdk/** and ai/** to workflow watched paths and compute-diff - Remove ai/* and api-reference/sdk/* from out-of-scope pre-filter - Add velt-reactions, velt-arrows, velt-area, velt-view-analytics, velt-node-sdk, velt-approval-engine to dispatch-updater-variant.sh - Bump Agent 2 max turns from 40 to 80 - Update mapping.md: add new mapped paths, remove stale unmapped entries, sync variant dispatch table Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Rakesh <rakesh@snippyly.com>
* v5.0.2-beta.29 * save --------- Co-authored-by: Rakesh <rakesh@snippyly.com>
* v5.0.2-beta.30 * docs: tighten 5.0.2-beta.30 changelog entries Rewrite the beta.30 New Features, Improvements, and Bug Fixes bullets into crisp, user-facing language and drop internal service/implementation details. Co-authored-by: Cursor <cursoragent@cursor.com> * save --------- Co-authored-by: Rakesh <rakesh@snippyly.com> Co-authored-by: Cursor <cursoragent@cursor.com>
* v5.0.2-beta.32 * docs: remove Comment Dialog Primitives section from api-methods API methods reference should document methods only, not primitive components. Removed the duplicated "Comment Dialog Primitives" section (90+ VeltCommentDialog* entries) from api-methods.mdx; these remain fully documented in comment-dialog/primitives.mdx (usage) and data-models.mdx (types). Updated three links that pointed to the removed #comment-dialog-primitives anchor. Co-authored-by: Cursor <cursoragent@cursor.com> * save --------- Co-authored-by: Rakesh <rakesh@snippyly.com> Co-authored-by: Cursor <cursoragent@cursor.com>
* v5.0.2-beta.33 * Bump Agent 1 max turns to 80 and pre-filter out-of-scope docs paths Agent 1 (docs-diff-planner) was hitting the 40-turn limit when processing large diffs. Bump to 80 and add shell-level pre-filtering in compute-diff.sh to drop paths that mapping.md marks as out-of-scope before the agent sees them. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * save * save --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Rakesh <rakesh@snippyly.com>
Co-authored-by: Rakesh <rakesh@snippyly.com>
…ges (#186) Bug 1: apply-patches.sh searched for changes.json inside $INCOMING_DIR (agent-skills-patches/) but artifacts extract changes.json to a sibling path (incoming/outputs/). Added $DOWNLOAD_ROOT to search the correct location, fixing 3 runs that produced patches but created no PR. Bug 2: data-models.mdx changes routed by "match type prefix to skill" but PermissionRequest had no matching prefix. Added an explicit model-family routing table to mapping.md so the planner can resolve types like PermissionRequest → velt-setup-best-practices. Bug 3: Pipeline reported success even when no PR was created. Added an assertion step that fails when tickets exist but no PR was produced, and an annotation step that warns on unmapped changes. Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
…ents customize behavior Add a new AI > Agent Comments page covering how AI agents can leave comments and findings via the Comment Annotations REST APIs, with explanation, examples, and accept/reject event handling. Also document agent comments and the suggestionAccepted/suggestionRejected events in the comments Customize Behavior reference, and wire the new page into navigation. Co-authored-by: Cursor <cursoragent@cursor.com>
* docs: add Agent Comments guide and document agent comments under comments customize behavior Add a new AI > Agent Comments page covering how AI agents can leave comments and findings via the Comment Annotations REST APIs, with explanation, examples, and accept/reject event handling. Also document agent comments and the suggestionAccepted/suggestionRejected events in the comments Customize Behavior reference, and wire the new page into navigation. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: link add/get comment annotation APIs and comment events from Agent Comments guide Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Cursor <cursoragent@cursor.com>
Agent 4 (skill-pr-composer) intermittently treats its prompt as informational context rather than instructions to execute, responding with "what would you like me to do?" instead of reading input files and creating a PR. This caused two consecutive pipeline failures. Prepend an explicit imperative instruction to the agent body so the model unambiguously treats it as a task to execute. Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* docs: add Agent Comments guide and document agent comments under comments customize behavior Add a new AI > Agent Comments page covering how AI agents can leave comments and findings via the Comment Annotations REST APIs, with explanation, examples, and accept/reject event handling. Also document agent comments and the suggestionAccepted/suggestionRejected events in the comments Customize Behavior reference, and wire the new page into navigation. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: link add/get comment annotation APIs and comment events from Agent Comments guide Co-authored-by: Cursor <cursoragent@cursor.com> * docs: add self-host data field inventory page and link from overview Move the exhaustive field inventory into its own Self-Host Data page, add it to the nav, and replace the duplicated inventory section in the overview with a link to the new reference. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Cursor <cursoragent@cursor.com>
* docs: add Agent Comments guide and document agent comments under comments customize behavior Add a new AI > Agent Comments page covering how AI agents can leave comments and findings via the Comment Annotations REST APIs, with explanation, examples, and accept/reject event handling. Also document agent comments and the suggestionAccepted/suggestionRejected events in the comments Customize Behavior reference, and wire the new page into navigation. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: link add/get comment annotation APIs and comment events from Agent Comments guide Co-authored-by: Cursor <cursoragent@cursor.com> * docs: add self-host data field inventory page and link from overview Move the exhaustive field inventory into its own Self-Host Data page, add it to the nav, and replace the duplicated inventory section in the overview with a link to the new reference. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: trim self-host field inventory intro and drop DB-engine names Remove the framing/omitted-by-design notes and the cross-cutting rules section, and replace Firestore/Firebase mentions with generic "Velt's DB". Co-authored-by: Cursor <cursoragent@cursor.com> * docs: remove built-in agent references from Agent Comments Drop built-in agent mentions and the agentType filter; the agent block now documents only custom (velt) and external agents. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: document the agent reason object fields Add a reason-object field table (with types, descriptions, examples), a suggestion vs suggestedFix note, and a fully-populated JSON example. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Cursor <cursoragent@cursor.com>
* docs: add Agent Comments guide and document agent comments under comments customize behavior Add a new AI > Agent Comments page covering how AI agents can leave comments and findings via the Comment Annotations REST APIs, with explanation, examples, and accept/reject event handling. Also document agent comments and the suggestionAccepted/suggestionRejected events in the comments Customize Behavior reference, and wire the new page into navigation. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: link add/get comment annotation APIs and comment events from Agent Comments guide Co-authored-by: Cursor <cursoragent@cursor.com> * docs: add self-host data field inventory page and link from overview Move the exhaustive field inventory into its own Self-Host Data page, add it to the nav, and replace the duplicated inventory section in the overview with a link to the new reference. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: trim self-host field inventory intro and drop DB-engine names Remove the framing/omitted-by-design notes and the cross-cutting rules section, and replace Firestore/Firebase mentions with generic "Velt's DB". Co-authored-by: Cursor <cursoragent@cursor.com> * docs: remove built-in agent references from Agent Comments Drop built-in agent mentions and the agentType filter; the agent block now documents only custom (velt) and external agents. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: document the agent reason object fields Add a reason-object field table (with types, descriptions, examples), a suggestion vs suggestedFix note, and a fully-populated JSON example. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: drop closing omitted-fields note from field inventory Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Cursor <cursoragent@cursor.com>
* docs: add Agent Comments guide and document agent comments under comments customize behavior Add a new AI > Agent Comments page covering how AI agents can leave comments and findings via the Comment Annotations REST APIs, with explanation, examples, and accept/reject event handling. Also document agent comments and the suggestionAccepted/suggestionRejected events in the comments Customize Behavior reference, and wire the new page into navigation. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: link add/get comment annotation APIs and comment events from Agent Comments guide Co-authored-by: Cursor <cursoragent@cursor.com> * docs: add self-host data field inventory page and link from overview Move the exhaustive field inventory into its own Self-Host Data page, add it to the nav, and replace the duplicated inventory section in the overview with a link to the new reference. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: trim self-host field inventory intro and drop DB-engine names Remove the framing/omitted-by-design notes and the cross-cutting rules section, and replace Firestore/Firebase mentions with generic "Velt's DB". Co-authored-by: Cursor <cursoragent@cursor.com> * docs: remove built-in agent references from Agent Comments Drop built-in agent mentions and the agentType filter; the agent block now documents only custom (velt) and external agents. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: document the agent reason object fields Add a reason-object field table (with types, descriptions, examples), a suggestion vs suggestedFix note, and a fully-populated JSON example. Co-authored-by: Cursor <cursoragent@cursor.com> * docs: drop closing omitted-fields note from field inventory Co-authored-by: Cursor <cursoragent@cursor.com> * save --------- Co-authored-by: Cursor <cursoragent@cursor.com>
* fix: correct VS Code install button URL on Docs MCP page * fix: update Docs MCP URLs to velt.dev/docs/mcp --------- Co-authored-by: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com>
…dd Memory overview Style and accuracy pass over PR #180 against the internal developer guides (approval-engine-developer-guide.md, memory-developer-guide.md): - Add ai/memory/overview.mdx: concept-first page (what Memory is, how it works pipeline, core concepts, data provenance, limitations) so a new web dev gets the what/why before the hands-on guide; wire into docs.json. - Slim ai/memory/developer-guide.mdx to the hands-on journey; concepts now live in the overview. Keep the full add-knowledge -> ask -> search -> suggest tutorial. - Remove em-dashes and en-dash ranges across all beta.31 pages; switch "We'll" to second person; convert bold Important-style asides to callouts. - Drop internal terminology from public docs: onRequest/onCall Cloud Function naming, Firestore collection name knowledge_chunks. - Changelog beta.31: reposition entry chronologically (merge with main had stranded it next to beta.22), remove backend-only bullets per changelog rules (Stripe billing fan-out, FieldValue sentinels, log-level change, Firestore path normalization), reframe remaining bullets to observable behavior, and link Memory bullets to the new overview/guide. - Move APPROVAL-ENGINE-MEMORY-DOC-GAPS.md from repo root to .claude/logs/ (build-ignored) and refresh its fixed/open status. https://claude.ai/code/session_015qRRK7w4nbVX64YSGJDWZn
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Style and accuracy pass over PR #180 against the internal developer guides. Merging this into the
v5.0.2-beta.31branch updates PR #180 directly.ai/memory/overview.mdx(concept-first: what Memory is, how it works, core concepts, data provenance, limitations) and wire it intodocs.json.knowledge_chunkscollection name)..claude/logs/(build-ignored).https://claude.ai/code/session_015qRRK7w4nbVX64YSGJDWZn
Generated by Claude Code