/ultrareview returns empty findings + consumes free credit on large PR (Issue #50029 repro)
Hitting the documented empty-findings bug from #50029 on PR
https://github.com/armaansarora/internship-command-center/pull/3
(301 files, 19,491 lines, 141 commits, all 5 CI checks green).
Two consecutive /ultrareview 3 runs both returned "Remote review completed"
with no findings and consumed free credits. Session IDs:
- session_01TWvuVw7khkrcQEPx1ox7AR
- session_01Jrg1fCvfaBprbDpfd8BxtR
The PR is well-formed (mergeable, clean merge state, no binary blobs in the diff,
no oversized lockfile changes beyond a normal package-lock update). The diff
profile matches the empirical danger zone documented in #50029 (>200 files,
≥10K-line changes). Both runs collapsed to empty findings silently — no error
message, no partial output, no actionable signal.
Requesting both credits be refunded per the precedent in #52780, #53455, and
#49706 since this matches the documented silent-failure mode on large diffs.
Happy to provide additional repro detail if useful.
Also flagging that the second attempt was made AFTER the first crashed — there's
no warning in the CLI confirm dialog that empty-findings is a known failure
mode for diffs of this size. A "warn-and-confirm" check would have prevented
me from burning the second credit on a known-bad-shape PR.
This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗