[BUG] /ultrareview returned empty findings ([]) on large-scope branch — free credit consumed
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
## Summary
/ultrareview returned empty findings ([]) on a large-scope branch review,
and the free credit was still decremented (counter advanced from "1 of 3" to "2 of 3").
This matches the pattern reported in #50029, #53010, #54516, and #54871.
## Environment
- Platform: macOS (Darwin 25.4.0)
- Shell: zsh
- Claude Code: CLI
- Model: Opus 4.7
## Repro
- On branch
review/phase-23-wide(target:main) - Run
/ultrareview(no args) - CLI reports:
Free ultrareview 2 of 3.Scope: 177 files changed, 22173 insertions(+), 8044 deletions(-)- Track URL issued, ~5–10 min cloud run
- Task notification arrives:
status: completed,summary: Remote review completed - Findings payload:
[]
## Session details
- Session ID:
session_01BeLYYxmoA8VTjVrfTPP5DM - Branch:
review/phase-23-wide→main - Diff size: 177 files, +22173 / −8044
- Result: empty findings array
- Credit state after run: 2 of 3 free runs consumed
What Should Happen?
## Expected
Either:
- (a) Findings returned for a diff of this size, OR
- (b) Free credit NOT decremented when the remote agent returns an empty payload
on a non-trivial diff (server-side failure indistinguishable from a clean codebase).
## Actual
Empty [] returned with no error surfaced to the user, free credit consumed.
Error Messages/Logs
Steps to Reproduce
This matches the pattern reported in #50029, #53010, #54516, and #54871.
## Related
- #50029, #53010, #54516, #54871, #49905
Claude Model
None
Is this a regression?
Yes, this worked in a previous version
Last Working Version
_No response_
Claude Code Version
2.1.128
Platform
Anthropic API
Operating System
macOS
Terminal/Shell
Terminal.app (macOS)
Additional Information
## Request
Please restore the consumed free credit for session
session_01BeLYYxmoA8VTjVrfTPP5DM, and consider gating the credit decrement
on a non-empty response (or a clear "clean codebase" signal distinct from
silent server-side failure).
## Related
- #50029, #53010, #54516, #54871, #49905
This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗