What
code-quality.yml's two informational mypy steps, and the strict autobot_shared
check's suggested suppression syntax, all cite GH#7105 as the tracker for their
pre-existing error counts:
code-quality.yml:500 — "Errors in the backends are tracked in GH#7105."
code-quality.yml:503 — suggests # type: ignore[code] # GH#7105 for new suppressions
code-quality.yml:510 — "4133 pre-existing errors tracked for resolution in GH#7105" (autobot-backend/, informational)
code-quality.yml:515 — "1281 SQLAlchemy Column[T] errors tracked in GH#7105" (autobot-slm-backend/, informational)
#7105 is CLOSED ("ci: mypy is installed but never invoked — silent-tuple bug
from #7078 would have been caught"). It tracked making mypy run at all, not
resolving what running it found. Once it closed, mypy started running and
surfaced 4133 + 1281 pre-existing errors — but nothing was filed to track
those, so the workflow kept citing the issue that was actually done.
Found while fixing #15333 AC3, which flagged this exact citation as stale.
Why this matters
A reader hitting the GH#7105 reference today finds a closed issue and
reasonably concludes the backlog is either resolved or being tracked
elsewhere. Neither is true — the two steps are still (informational) /
non-blocking, and the counts in the comments (4133, 1281) are whatever was
true when someone last hand-updated them, not a live measurement.
Scope
This issue is the tracking placeholder the workflow comments should cite
instead of #7105. It does not itself fix the underlying mypy errors — that
is 4133 + 1281 individual findings, almost certainly wrong to batch into one
ticket. Whoever picks this up should:
- Re-measure the current error counts for
autobot-backend/ and
autobot-slm-backend/ (the numbers in the workflow comments are unverified
as of this filing).
- Decide whether the backlog should be broken into per-package or
per-error-code follow-up issues, or resolved as one sustained effort.
- Update or retire the
code-quality.yml comments once a decision is made,
the same way #15333 retired the #7105 citation.
Acceptance criteria
What
code-quality.yml's two informational mypy steps, and the strictautobot_sharedcheck's suggested suppression syntax, all cite
GH#7105as the tracker for theirpre-existing error counts:
code-quality.yml:500— "Errors in the backends are tracked in GH#7105."code-quality.yml:503— suggests# type: ignore[code] # GH#7105for new suppressionscode-quality.yml:510— "4133 pre-existing errors tracked for resolution in GH#7105" (autobot-backend/, informational)code-quality.yml:515— "1281 SQLAlchemy Column[T] errors tracked in GH#7105" (autobot-slm-backend/, informational)#7105is CLOSED ("ci: mypy is installed but never invoked — silent-tuple bugfrom #7078 would have been caught"). It tracked making mypy run at all, not
resolving what running it found. Once it closed, mypy started running and
surfaced 4133 + 1281 pre-existing errors — but nothing was filed to track
those, so the workflow kept citing the issue that was actually done.
Found while fixing #15333 AC3, which flagged this exact citation as stale.
Why this matters
A reader hitting the
GH#7105reference today finds a closed issue andreasonably concludes the backlog is either resolved or being tracked
elsewhere. Neither is true — the two steps are still
(informational)/non-blocking, and the counts in the comments (4133, 1281) are whatever was
true when someone last hand-updated them, not a live measurement.
Scope
This issue is the tracking placeholder the workflow comments should cite
instead of
#7105. It does not itself fix the underlying mypy errors — thatis 4133 + 1281 individual findings, almost certainly wrong to batch into one
ticket. Whoever picks this up should:
autobot-backend/andautobot-slm-backend/(the numbers in the workflow comments are unverifiedas of this filing).
per-error-code follow-up issues, or resolved as one sustained effort.
code-quality.ymlcomments once a decision is made,the same way
#15333retired the#7105citation.Acceptance criteria
autobot-backend/andautobot-slm-backend/are re-measured and recorded herecode-quality.yml's suppression comments cite whichever issue(s) actually track the resolution work