Skip to content

chore(ci): add fix-markers for recent merges (#559, #561, #588, #593, #590-CI)#606

Merged
ruvnet merged 1 commit into
mainfrom
chore/fix-markers-for-recent-merges
May 17, 2026
Merged

chore(ci): add fix-markers for recent merges (#559, #561, #588, #593, #590-CI)#606
ruvnet merged 1 commit into
mainfrom
chore/fix-markers-for-recent-merges

Conversation

@ruvnet

@ruvnet ruvnet commented May 17, 2026

Copy link
Copy Markdown
Owner

Summary

Adds 6 new entries to `scripts/fix-markers.json` so the regression guard catches a future revert of any fix that just landed via PRs #590, #584, #595, #604, #602:

Marker ID What it asserts
`RuView#559` `./verify` PROOF_DIR/V1_SRC point at `archive/v1/` (post-archive layout)
`RuView#561` README has the correct flash offsets (app @ 0x20000, ota_data @ 0xf000) AND points at `firmware/esp32-csi-node/provision.py`; FORBIDS the old `0x10000 …esp32-csi-node.bin` and `python scripts/provision.py` strings
`RuView#588-SEC020` Both `provision.py` files print `(set)`/`(empty)`, not `'*' * len(args.password)`; FORBIDS the asterisk-run pattern
`RuView#593` `vital_signs.rs` exports `phase_circular_variance`, has the standard circular-variance comment, and ships `test_phase_variance_handles_wraparound`
`RuView#590-fuzz-stub` `esp_stubs.h` declares `wifi_ps_type_t` / `WIFI_PS_NONE` / `esp_wifi_set_ps` (keeps Fuzz Testing job green)
`RuView#590-swarm-test` `scripts/qemu_swarm.py` passes `--force-partial` (keeps Swarm Test ADR-062 job green)

Each marker carries a `rationale` linking to the original incident, so a future contributor who hits a guard failure understands why the assertion is there.

Verification

```
$ python scripts/check_fix_markers.py
Fix-marker regression guard - 17 marker(s)

All 17 fix markers present.
```

CI's `Fix-Marker Regression Guard` workflow will run the same check on every PR.

Why this matters

Three of these fixes shipped on PR #590 (#559, #561, #588) and one on PR #595 (#593). The other two cover the CI infrastructure fixes that brought the Fuzz Testing and Swarm Test (ADR-062) jobs back to green after ~5 weeks red on `main`. Without these markers a future cleanup-PR could trivially revert any of them in a refactor without any signal at review time.

🤖 Generated with claude-flow

…590-CI)

Six new entries in scripts/fix-markers.json so the regression guard
(.github/workflows/fix-regression-guard.yml + scripts/check_fix_markers.py)
catches a future revert of any of these fixes:

- RuView#559 — ./verify points at archive/v1/ paths
- RuView#561 — README app flash offset 0x20000 + ota_data_initial.bin at 0xf000
                + canonical provision.py path
- RuView#588-SEC020 — provision.py prints (set)/(empty), not '*' * len(pw)
                (forbids the asterisk-run pattern that leaks password length)
- RuView#593 — vital_signs.rs uses phase_circular_variance for wrapped phases
- RuView#590-fuzz-stub — esp_stubs.h declares wifi_ps_type_t / WIFI_PS_NONE
                / esp_wifi_set_ps (keeps Fuzz Testing job green)
- RuView#590-swarm-test — qemu_swarm.py passes --force-partial to provision.py
                (keeps Swarm Test ADR-062 job green)

Verified: `python scripts/check_fix_markers.py` reports All 17 fix markers
present.

Co-Authored-By: claude-flow <ruv@ruv.net>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant