feat(DRC-2641): run history view + metadata Activity runs#1242
Open
feat(DRC-2641): run history view + metadata Activity runs#1242
Conversation
iamcxa
changed the title
Contributor
There was a problem hiding this comment.
Pull request overview
Adds run attribution and a run-history/timeline experience across backend + MCP + UI so users can see who triggered runs (user vs Recce AI) and include metadata-only checks (schema/lineage diff) in Activity.
Changes:
- Introduces
Run.triggered_byand propagates it through REST + MCP tool schemas andsubmit_run(). - Persists “metadata runs” for
schema_diff/lineage_diffso they appear in Activity, and updates run name generation for those types. - Updates UI to display actor badges and merges run history into the check timeline (with new component/tests).
Reviewed changes
Copilot reviewed 22 out of 22 changed files in this pull request and generated 5 comments.
Show a summary per file
| File | Description |
|---|---|
| tests/test_mcp_server.py | Updates MCP server tests for triggered_by propagation and metadata-run persistence behavior. |
| tests/apis/test_run_func.py | Adds unit tests for triggered_by propagation/defaulting in submit_run(). |
| recce/util/recce_cloud.py | Improves RecceCloudException.__str__ formatting for clearer logs/errors. |
| recce/state/cloud.py | Exports checks (not just runs) when uploading state to session in cloud mode. |
| recce/models/types.py | Adds Run.triggered_by field to the Run model. |
| recce/models/check.py | Implements cloud-mode “write-through” so created checks are also kept in local state. |
| recce/mcp_server.py | Adds metadata-run creation helper; extends MCP schemas and ensures run records exist for all check types. |
| recce/apis/run_func.py | Adds lineage/schema run-name generation and triggered_by support in submit_run(). |
| recce/apis/run_api.py | Hardcodes REST-triggered runs as triggered_by="user"; includes error + triggered_by in run listing. |
| recce/apis/check_api.py | Hardcodes REST-triggered reruns as triggered_by="user". |
| js/packages/ui/src/components/check/timeline/index.ts | Re-exports RunTimelineEntry and its types. |
| js/packages/ui/src/components/check/timeline/tests/RunTimelineEntry.test.tsx | Adds unit tests for RunTimelineEntry. |
| js/packages/ui/src/components/check/timeline/tests/CheckTimelineOss.merge.test.tsx | Adds tests for merging/sorting timeline events + runs. |
| js/packages/ui/src/components/check/timeline/RunTimelineEntry.tsx | New UI component for run items within the check timeline. |
| js/packages/ui/src/components/check/timeline/CheckTimelineOss.tsx | Fetches runs (cloud-only) and merges them into the activity timeline; derives display status. |
| js/packages/ui/src/components/check/tests/CheckCard.actorBadge.test.tsx | Adds tests for check actor badge rendering behavior. |
| js/packages/ui/src/components/check/CheckListOss.tsx | Wires actor_type from API response into CheckCard props. |
| js/packages/ui/src/components/check/CheckDetailOss.tsx | Displays actor badge on check detail header (skipping preset). |
| js/packages/ui/src/components/check/CheckCard.tsx | Adds actor badge support and exports shared badge config. |
| js/packages/ui/src/api/types/run.ts | Adds triggered_by to run API type. |
| js/packages/ui/src/api/checks.ts | Adds cloud-only actor_type to check API type. |
Comment on lines
+126
to
+142
| // Fetch runs only in cloudMode (authToken present = Cloud) | ||
| const { data: checkRuns } = useQuery({ | ||
| queryKey: ["check-runs", checkId], | ||
| queryFn: async () => { | ||
| const allRuns = await listRuns(apiClient); | ||
| return allRuns | ||
| .filter((r) => r.check_id === checkId) | ||
| .map( | ||
| (r): RunEntry => ({ | ||
| run_id: r.run_id, | ||
| run_at: r.run_at, | ||
| status: deriveRunStatus(r.status, r.error), | ||
| summary: r.error || undefined, | ||
| triggered_by: r.triggered_by, | ||
| }), | ||
| ); | ||
| }, |
There was a problem hiding this comment.
CheckTimelineOss fetches all runs via listRuns() and then filters client-side by check_id. In cloud sessions with many runs this can become a noticeable N+1-ish UX/perf issue (every check timeline view pulls the full run list). Prefer adding an API filter (e.g. /api/runs?check_id=...) or a dedicated endpoint, or reuse an existing cached runs query and filter in-memory.
Contributor
Author
There was a problem hiding this comment.
Valid concern. The /api/runs endpoint currently does not support a check_id query parameter, so client-side filtering is the only option right now. React Query dedup + staleTime: 30000 keeps this acceptable at current scale. Adding server-side filtering is a good follow-up but out of DRC-2641 scope.
js/packages/ui/src/components/check/timeline/RunTimelineEntry.tsx
Outdated
Show resolved
Hide resolved
Comment on lines
+241
to
+247
| # Write-through: also keep in local state so that: | ||
| # 1. Runs created after this check can reference it | ||
| # 2. export_persistent_state() includes checks in recce_state.json | ||
| # 3. Preview instance gets checks + runs together from S3 sync | ||
| self._checks.append(new_check) | ||
|
|
||
| logger.debug(f"Created check {new_check.check_id} in cloud and local state") |
There was a problem hiding this comment.
In cloud mode, create() now appends the created check to local context.checks (write-through). However update_check_by_id() and delete() in cloud mode don’t update/remove the corresponding entry from context.checks, so subsequent export_state() / state sync can include stale or deleted checks. Consider updating cloud-mode update/delete paths to also reconcile self._checks (replace by check_id / remove by check_id) to keep local state consistent.
Contributor
Author
There was a problem hiding this comment.
The write-through on create() was specifically added so the MCP agent can reference newly-created checks for subsequent runs and export_state(). In cloud mode, update_check_by_id and delete delegate to the cloud API (source of truth) and do not need local reconciliation — find_check_by_id also reads from cloud API directly. Adding local sync to update/delete would risk race conditions between local cache and cloud state. Keeping this asymmetry intentional for now.
Codecov Report❌ Patch coverage is
... and 1 file with indirect coverage changes 🚀 New features to boost your workflow:
|
Contributor
Author
Signed-off-by: Kent <iamcxa@gmail.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
Signed-off-by: Kent <iamcxa@gmail.com>
…641) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
Timeline self-fetches runs in cloudMode using useApiConfig. No props threading needed from Cloud wrapper. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
- Add triggered_by field to Run model ("user" | "recce_ai")
- submit_run accepts triggered_by parameter
- HTTP handlers hardcode triggered_by="user" to prevent spoofing
- generate_run_name supports lineage_diff and schema_diff types
- Run list response includes error field
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Kent <iamcxa@gmail.com>
- _create_metadata_run helper creates Run records for schema_diff and lineage_diff without going through submit_run (no task handler) - run_check now creates Run for metadata-only types instead of short-circuiting (so they appear in Activity) - create_check auto-runs metadata-only types via manifest read - Add triggered_by to run_check and create_check tool schemas Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
- Skip actor badge for preset checks in CheckCard and CheckDetailOss - deriveRunStatus: finished+error=error, finished+no error=success - RunTimelineEntry shows "by AI" label for recce_ai triggered runs - BaseRun type includes triggered_by field - Update CheckCard actor badge tests for preset skip Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
Previously _export_state_to_session set checks=[] to avoid conflicts with Cloud DB. Now includes agent-created checks so Cloud can import them into the DB for PR comment checklist display. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
- submit_run triggered_by propagation (2 tests) - _create_metadata_run helper (3 tests) - create_check/run_check triggered_by flow (5 tests) - Fix 5 existing tests broken by _create_metadata_run + triggered_by changes Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
…RC-2641)
RecceCloudException.__str__ now returns "[HTTP {code}] {reason}" instead
of just the message. This makes API failures debuggable in Langfuse traces
where only str(e) is captured.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Kent <iamcxa@gmail.com>
…nc (DRC-2641) In cloud mode, CheckDAO.create() now appends the check to local state after writing to Cloud API. This ensures: 1. Runs created after the check can reference it locally 2. export_persistent_state() includes checks in recce_state.json 3. Preview instance gets checks + runs together from S3 sync Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
…g, tool description (DRC-2641)
- Fix deriveRunStatus case mismatch: use toLowerCase() to match RunStatus enum
("Finished"/"Failed" not "finished"/"failed") — run badges now show correct colors
- Add try/except to metadata run path in create_check — ensures export_persistent_state
runs even when lineage_diff/schema_diff raises
- Update run_check tool description to document Run object response shape
- Add "Failed" → "error" mapping to deriveRunStatus
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Kent <iamcxa@gmail.com>
…r (DRC-2641) _export_state_to_session now copies checks (line 427 in cloud.py), but the test still asserted checks==[]. Update to match actual behavior. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
…ing, typing - Fix test_mcp_e2e.py: update lineage_diff/schema_diff assertions to match Run object return structure (result nested in result["result"]) - RunTimelineEntry: render as semantic <button> with aria-label when onClick is provided for keyboard accessibility - mcp_server.py: wrap metadata branch (lineage/schema diff) in try/except RecceException for consistent error shape - types.py: narrow triggered_by from Optional[str] to Optional[Literal["user", "recce_ai"]] for model-level validation Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
… path - Add 4 tests for generate_run_name with lineage_diff and schema_diff types (no params, single node, multiple nodes) - Add test for RecceException in metadata branch of _tool_run_check (mirrors existing query-based branch error test) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Kent <iamcxa@gmail.com>
iamcxa
force-pushed
the
feature/drc-2641-run-history-view
branch
from
c16588a to
8548464
Compare
Contributor
Author
MCP E2E Test Results (DRC-2641)All 6 MCP E2E tests pass after fixing
Recording
What was fixed
|
…ck_id passthrough
- generate_run_name: support node_id (REST), node_names (MCP array),
and node_ids (MCP array) for schema_diff naming. Extract short name
from fully-qualified IDs via split('.')[-1]
- create_run_handler: pass check_id=input.check_id to submit_run()
so REST-created runs preserve the check association
- Tests: 7 cases covering all three param conventions
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Kent <iamcxa@gmail.com>
Same 3-convention pattern as schema_diff: node_names (MCP array) then
node_ids (MCP fully-qualified IDs). Extracts short name via split('.')[-1].
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: Kent <iamcxa@gmail.com>
iamcxa
commented
Mar 27, 2026
Contributor
Author
iamcxa
left a comment
iamcxa
left a comment
There was a problem hiding this comment.
Self-review after addressing all Copilot feedback + code-reviewer agent findings.
Fixes applied (4 commits)
| Commit | Scope |
|---|---|
| a838a33 | Test assertions for Run object return, RunTimelineEntry a11y, metadata error handling, Literal type |
| c16588a | Coverage: generate_run_name + metadata error path tests |
| 30deecc | Schema param convention (node_id/node_names/node_ids), check_id passthrough in create_run_handler |
| 43347a1 | row_count_diff node_ids fallback (same 3-convention pattern) |
Review agent findings
| # | Severity | Verdict |
|---|---|---|
| C1 (metadata error catch) | NIT | Outer call_tool catches all Exception — no actual gap |
| C2 (update return discarded) | NIT | Cloud API updates in-place, UUID never changes |
| H1 (row_count_diff node_ids) | MEDIUM | Fixed in 43347a1 |
| H2 (state.checks reset) | PRE-EXIST | Line 258 unchanged by PR — cloud mode fetches from API |
| H3 (CheckDAO self vs new) | PRE-EXIST | Local mode pre-existing |
Pre-scan
- Zero CLAUDE.md violations
- Zero stale references
- MCP tool description matches actual return format
Remaining deferred (2 threads)
- CheckTimelineOss N+1 runs fetch — needs server-side filter endpoint
- check.py write-through asymmetry — intentional, cloud API is source of truth
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
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.
Summary
triggered_byfield ("user" | "recce_ai") for run attribution_create_metadata_runcreates Activity entries for schema_diff/lineage_diffrun_checkandcreate_checknow produce Run records for ALL check typestriggered_byparameter added torun_checkandcreate_checkderiveRunStatusmaps finished+error to error statusKey changes
recce/models/types.pyRun.triggered_byfieldrecce/apis/run_func.pysubmit_runacceptstriggered_by, name gen for lineage/schemarecce/apis/run_api.pyrecce/apis/check_api.pytriggered_by="user"recce/mcp_server.py_create_metadata_run, metadata runs in run_check/create_checkDepends on
Test plan
🤖 Generated with Claude Code