Repository navigation
fix(resume): capture pause block names and keep the resume page off the execution snapshot - #10104
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
[Medium impact] The PR appears safe to merge; no actionable new issues were found. SummaryThe resume page no longer downloads or repeatedly parses the execution snapshot to show block names.
|
… display reads off the snapshot
|
@cubic-dev-ai review this PR |
@waleedlatif1 Sorry, I encountered an error while trying to start the review. Please try again later. |
|
@cubic-dev-ai review this PR |
@waleedlatif1 I have started the AI code review. It will take a few minutes to complete. |
There was a problem hiding this comment.
All reported issues were addressed across 22 files
Reply with feedback, questions, or to request a fix.
Turn on auto-fix | Re-trigger cubic
Summary
JSON.parsed it once per pause point on every render (including every keystroke) just to read a block name. Large runs with many pause points made the page take minutes to load.blockNameon each pause point when the pause is captured, and persistence stores it inpause_pointspause-block-names.tsnames pause points for every reader (resume page reads and the v2 run read): the captured name first, else the run's deployment version, read inside the DB and only when a point lacks a name. The snapshot is never parsed for names.getPausedExecutionViewreads without the snapshot column.triggerIdscome frommetadata(already written at pause time), so Postgres only detoasts the snapshot for rows saved before that.getPausedExecutionDetailis the view plus the snapshot, for resume and API callers.readPausedWorkflowExecutiontakes a requiredincludeSnapshot.GET /api/resume/{workflowId}/{executionId}?includeSnapshot=defaults totrue, so the documented API response is unchanged; the resume page sendsfalse. The/api/workflows/{id}/paused/{executionId}API route always includes it.executionno longer carries the snapshot, which matches its documentedPausedExecutionSummaryshapeblockNameis now set for manual runs too, andblockIdstrips subflow clone suffixes (stripCloneSuffixes). The v2 contract description and OpenAPI are updated, andopenapi-core.jsondocumentsincludeSnapshotandPausePoint.blockName.Type of Change
Testing
pause-persistence.integration.tsagainst real Postgres:readPausedRunPointson a manual runincludeSnapshotreturns ittriggerIdscome from metadata, falling back to the snapshotengine.test.ts: the paused result carries the DAG node's block name (red when the capture is removed)api/resume/[workflowId]/[executionId]/route.test.ts: onlyincludeSnapshot=falseopts out; the default and the paused route read the snapshot (red when the flag is ignored)bun run lint; apptype-check;check:openapi,check:api-validation:strict,check:explicit-any(baseline shrunk),check:test-patterns,check:boundaries,check:client-boundary,check:source-conventions,check:file-names,check:utils,check:test-types; block-registry checkChecklist
test-auditauthoring gate)🤖 Generated with Claude Code