CI: give scope-stats the DFX smoke step the other four have

[ChaoWao]

Summary

tests/st/{a2a3,a5}/tensormap_and_ringbuffer/dfx/scope_stats/ is the only test of the scope-stats capture pipeline, and its artifact validation never ran in any pipeline.

_validate_scope_stats_artifact returns early unless --enable-scope-stats is on (test_scope_stats.py:117), and nothing passed it:

$ grep -rn "scope_stats\|enable-scope-stats" .github/
$ # empty

$ # meanwhile, in the same two workflows:
      3 enable-dep-gen
      1 enable-chip-swimlane
      1 enable-pmu
      1 --dump-args

To be precise about what was and wasn't covered: the scene half did run. The class is an ordinary SceneTestCase, so the main sweep compiled and ran its vector_example and checked the golden. What never ran is the part that matters for this feature — the scope_stats.jsonl assertions: at least four begin/end records, dropped == 0, the metadata schema, and the required per-record fields. A collector that silently emitted nothing, or dropped every record, would have kept CI green.

Both new steps copy the shape of the args_dump step beside them, including the manual_mode mapping that keeps only as only and otherwise uses include, since these steps own the whole corpus under their path.

docs/ci.md had recorded the omission as a fact — "Scope-stats has no dedicated CI smoke: its ordinary scene test stays in the main sweep, and artifact validation runs only when --enable-scope-stats is supplied explicitly." It now describes the step, and why scope-stats needed one for a different reason than the others: for the other four the dedicated step exists so coverage survives a case being marked manual, whereas here the scene test runs either way and only the artifact assertions were gated off.

Testing

• check-yaml equivalent (yaml.safe_load) on both workflows; markdownlint-cli2 clean on docs/ci.md

• Both new steps run exactly as the workflows invoke them, including --manual include:

  $ pytest tests/st/a2a3/tensormap_and_ringbuffer/dfx/scope_stats/ \
      --platform a2a3sim --device 0-15 -p no:xdist --pto-session-timeout 600 \
      --require-pto-isa --manual include --enable-scope-stats
  1 passed
  
  $ pytest tests/st/a5/tensormap_and_ringbuffer/dfx/scope_stats/ \
      --platform a5sim --device 0-15 -p no:xdist --pto-session-timeout 600 \
      --require-pto-isa --manual include --enable-scope-stats
  1 passed

• The assertions are not vacuous — the validator checks record count, dropped == 0, schema and per-record fields, so the step fails if the collector stops producing.

This touches .github/workflows/_*.yml, which .claude/rules/ci-change-detection.md §2 keeps out of NON_CODE precisely so a CI-implementation change runs the whole matrix — including the two jobs this PR modifies.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 79bd9e40-1f15-4726-acb2-6c2ef55e3bb7

📥 Commits

Reviewing files that changed from the base of the PR and between 66ba5c4 and 57285a6.

📒 Files selected for processing (3)

• .github/workflows/_st-sim-a2a3.yml
• .github/workflows/_st-sim-a5.yml
• docs/ci.md

Included review availability: Your plan provides up to 1 included review per hour; 0 remain after this review.

---

📝 Walkthrough

Walkthrough

The CI workflows add dedicated scope-stats smoke steps for A2A3 and A5. The CI documentation describes their mode selection and scope_stats.jsonl artifact validation.

Changes

Scope-stats CI coverage

Layer / File(s)	Summary
Dedicated scope-stats smoke execution .github/workflows/_st-sim-a2a3.yml, .github/workflows/_st-sim-a5.yml, docs/ci.md	The A2A3 and A5 workflows run the scope-stats suites when include_dfx_smokes is enabled. The steps enable scope-stat collection and preserve the manual_mode selection. The documentation describes scope_stats.jsonl validation.

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to 57285

This adds the missing scope-stats artifact checks to the two relevant CI workflows and documents the coverage change; no actionable merge-blocking risk remains after normal checks and review.

Poem

I’m a rabbit in the CI lane,
Scope stats hop through smoke-test rain.
A2A3 and A5 run bright,
JSONL checks confirm the flight.
“Only” or “include”—both paths are right!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the CI change: adding the missing DFX smoke step for scope-stats.
Description check	✅ Passed	The description directly explains the missing artifact validation, workflow changes, documentation update, and testing performed.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0 files. (3 skipped: 3 unsupported.)
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}