Skip to content

Commit c3bf2c2

Browse files
authored
Merge pull request #1957 from lidge-jun/codex/wave5-wp9-closeout
fix(antigravity): send a bare ide_version + Wave 5 closeout record
2 parents 5c66ad2 + 5d4151e commit c3bf2c2

2 files changed

Lines changed: 61 additions & 5 deletions

File tree

‎devlog/_plan/260817_wave5_execution/080_wave5d_antigravity.md‎

Lines changed: 32 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -185,3 +185,35 @@ make that record false rather than merely skip a step. Reported, not cleared.
185185

186186
The planned order (`#1889 → #1891 → #1897`) is therefore moot: #1897 is in, and the remaining
187187
two are gated on a human decision each — one a readiness checklist, one a security review.
188+
### Corrections from the WP8 audit
189+
190+
**Failing-check count.** #1889 has **two** distinct failing checks, `hygiene` and
191+
`enforce-target`. Earlier text said four, which was the count of failing check *runs* across
192+
re-runs (`enforce-target` appears three times). Verified with `unique`.
193+
194+
**#1891's head moved, and the reviewer's staleness finding is itself stale.** The audit reported
195+
the head 62 commits behind `origin/dev`, which would have mattered:
196+
`READINESS_LATEST_DEV_BEHIND_MAX = 10` in `.github/scripts/pr-quality-state.cjs` unticks the
197+
`latest_dev` box past that, so ticking without rebasing would have re-drafted the PR. Re-checked
198+
against the live head `81236807f`: **0 commits behind**. The author rebased in the interim, so
199+
ticking alone is now sufficient — which is what my comment on the PR says.
200+
201+
Worth keeping as a lesson rather than deleting: a rejected finding was still worth chasing,
202+
because the mechanism it named is real and would have made my advice wrong on a different day.
203+
204+
### Work found and done instead of held
205+
206+
The audit asked whether anything here could be landed rather than recorded. One thing could,
207+
and it was a live defect on `dev` independent of both PRs: `metadata.ide_version` in
208+
`src/oauth/google-antigravity.ts` was set to `antigravityUserAgent()` — the whole header,
209+
`antigravity/ide/2.5.5 (aidev_client; os_type=...; arch=...)` — where the real client sends
210+
`2.5.5`.
211+
212+
Nothing failed, which is why it survived: the request succeeds, it just does not look like
213+
Antigravity. `ANTIGRAVITY_IDE_VERSION` already existed one import away. Fixed in **#1955**, with
214+
a regression that pins the field and asserts the shapes it must not have; driven red first.
215+
216+
That is also the honest answer to "is the sponsorship refusal over-cautious": I hold #1889
217+
because reviewing *someone else's* auth change is the maintainer act the label records — but a
218+
one-line auth fix I wrote and verified myself is exactly the case where a maintainer sponsors
219+
their own work, so it ships.

‎devlog/_plan/260817_wave5_execution/090_wave6_closeout.md‎

Lines changed: 29 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -88,11 +88,16 @@ Run on the promotion candidate (local `dev`, 6 commits ahead of `origin/dev` at
8888

8989
### What actually closed, under the close-on-dev-merge decision
9090

91-
| Closed | Landed via |
92-
|--------|-----------|
93-
| #1894 | #1739 through PR #1921 |
94-
| #1843 | #1860, already released in v2.24.0 |
95-
| #1899 | superseded by the ordering assertion in PR #1923 |
91+
**Two issues** closed, plus one pull request:
92+
93+
| Closed | Kind | Landed via |
94+
|--------|------|-----------|
95+
| #1894 | issue | #1739 through PR #1921 |
96+
| #1843 | issue | #1860, already released in v2.24.0 |
97+
| #1899 | **pull request** | superseded by the ordering assertion in PR #1923 |
98+
99+
The first version of this table listed all three as issues, which overstated the run.
100+
#1899 is a PR; two issues closed, not three.
96101

97102
Everything else stayed open, and none of it for release-timing reasons — which is the point
98103
worth making about the policy change. It removed a gate that was never what held these back.
@@ -104,3 +109,22 @@ worth making about the policy change. It removed a gate that was never what held
104109
and `aca3c0241` were both cancelled by supersession as later merges landed. The local full
105110
suite above is the evidence that exists; a hosted run on the exact promotion head is the
106111
evidence that does not, and promotion should carry that distinction rather than bury it.
112+
### One merge landed on a red run
113+
114+
PR #1921's merge commit `9dbc5fc42` has a failing hosted run (`32026536154`). The failure is
115+
`provider request pacing queue > spaces concurrent starts in one provider FIFO` in
116+
`tests/request-pacing.test.ts` — a wall-clock assertion, which is the classic flake shape on a
117+
loaded macOS runner. Evidence it is not a live regression: the file passes locally, and every
118+
subsequent hosted run on `dev` is green including the current head.
119+
120+
It is recorded here because it happened, not because it blocks anything. A campaign record that
121+
omits the one merge that landed red is exactly the kind of record you cannot trust later.
122+
123+
### Promotion evidence, updated
124+
125+
The "no completed green run" statement above is **stale and superseded**. Run `32090176020` on
126+
`9eb3a101a` is `completed/success` with every job green — four test shards, macOS, keyring on
127+
all three OSes, npm-global on all three, gates, storage policy, api usage.
128+
129+
So the hosted evidence now exists. Promote the head CI actually evaluated; promoting a local ref
130+
that no run has seen would re-open the exact gap this section was written about.

0 commit comments

Comments
 (0)