fix(snippets): an annotated record is not a diverged one — a standing verdict vouches at its commit (#2782)
CI & Build / Python lint (push) Successful in 3s
CI & Build / Plugin hooks (push) Successful in 11s
CI & Build / TypeScript typecheck (push) Successful in 34s
CI & Build / integration (push) Successful in 29s
CI & Build / Python tests (push) Successful in 1m12s
CI & Build / Build & push image (push) Successful in 25s

Pull-time freshness confirms a cached body by containment: normalised cached
code must appear inside the fetched file. That is right for a record kept
verbatim and permanently wrong for a deliberately annotated one. A record
whose job is to say WHY the shape is what it is carries commentary the source
does not, so containment fails on every pull, forever — #2508 was reading
`diverged` although its declarations match the source exactly, and always
would. Annotation is a sanctioned record style, so this was two deliberate
designs colliding, and it was quietly poisoning the one honest drift signal:
decision #2707's scoreboard watches body_freshness for `diverged` and was
accruing false positives it could never age out.

The escape hatch is the verdict itself. verify_snippet is exactly where a
human or agent already judged this body a faithful rendering of that source,
and `verification.commit_sha` records the repo commit they judged it at — a
field whose own docstring (#2688) anticipated this: "makes 'the REPO moved on
since the check' computable, once the forge integration can compare it against
the current head." This is that comparison. When containment fails, a standing
verdict can still vouch, on four conditions and no fewer:

  - the verdict says `ok`;
  - it has not EXPIRED — verification_view recomputes code_sha against the
    record's current body, so editing the record retires the verdict;
  - it was not INVALIDATED by a push touching the location (#2691);
  - the file just fetched is at the very commit the verdict was stamped at.

That last one is what keeps it honest: the hatch vouches for a body against
ONE known commit, never against whatever the source became since. The moment
the file moves, containment resumes as the authority and the record reads
`diverged` until someone re-verifies — correct, because at that point nobody
has looked. The first three are checked by reusing verification_view rather
than restating its rule, so "expired" keeps meaning one thing.

Nothing is rewritten and no new freshness value is minted; `verification`
already travels in the same payload, so a reader can see the basis rather
than take "current" on faith. Verdicts predating commit stamping carry no
commit to compare and therefore do not vouch — they fall through to
containment rather than passing on age alone.

Tests pin the fix and, more usefully, every condition that switches it back
off: moved commit, expired verdict, push-invalidated verdict, non-ok verdict,
no verdict, and a pre-#2688 verdict with no commit_sha. Plus a regression
that a verbatim record still takes the containment path untouched.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
2026-08-23 22:36:04 -04:00
co-authored by Claude Fable 5
parent 6871c25445
commit 446d6da0d7
3 changed files with 141 additions and 0 deletions
+48
View File
@@ -987,6 +987,46 @@ async def _refresh_provenance(note, commit_sha: str) -> None:
await notes_svc.update_note(note.user_id, note.id, data=data)
def _verdict_still_vouches(note, fields: dict, fetched_commit_sha: str) -> bool:
"""Does a standing `ok` verdict still speak for this body, at this commit?
Containment (cached code ∈ fetched file) is the fast path, and it is right
for a record kept verbatim. It is WRONG for a deliberately annotated one
(#2782): a record whose job is to say why the shape is what it is carries
commentary the source does not, so containment fails forever and the record
reads `diverged` on every pull. That turns the one honest drift signal into
a permanent false positive — and annotation is a sanctioned record style,
so this is two deliberate designs colliding, not a malformed record.
The escape hatch is the verdict itself. `verify_snippet` is precisely where
a human or agent already judged this body a faithful rendering of that
source, and `verification.commit_sha` records the repo commit they judged
it at — a field whose own docstring (#2688) anticipated this use: "makes
'the REPO moved on since the check' computable, once the forge integration
can compare it against the current head." This is that comparison.
All four conditions, and none is optional:
- the verdict says `ok`;
- it has not EXPIRED — `verification_view` recomputes `code_sha` against
the record's current body, so editing the record retires the verdict;
- it was not INVALIDATED by a push touching the location (#2691);
- the file we just fetched is at the very commit the verdict was stamped
at. Any later commit means nobody has judged what is there now.
The last one is what keeps this honest: it vouches for a body against ONE
known commit, never against whatever the source has become since. The
moment the file moves, containment resumes as the authority and the record
reads `diverged` until someone re-verifies — which is the correct outcome,
because at that point nobody has looked.
"""
if not fetched_commit_sha:
return False
view = verification_view(note, fields)
if view.get("status") != VERIFY_OK or view.get("needs_attention"):
return False
return view.get("commit_sha") == fetched_commit_sha
async def attach_live_body(note, data: dict) -> None:
"""Decorate a PULL response with forge-checked freshness (#2690).
@@ -1115,6 +1155,14 @@ async def attach_live_body(note, data: dict) -> None:
_refresh_provenance(note, fetched.commit_sha),
site="pull provenance-refresh",
)
elif _verdict_still_vouches(note, fields, fetched.commit_sha or ""):
# Containment failed, but an unexpired `ok` verdict stamped at exactly
# this commit already judged this body a faithful rendering of it —
# the annotated-record case (#2782). Trust the judgment over the
# substring test; `data["verification"]` travels in the same payload,
# so a reader can see the basis rather than take "current" on faith.
data["body_source"] = "forge"
data["body_freshness"] = "current"
else:
data["body_source"] = "cache"
data["body_freshness"] = "diverged"