Skip to content

feat(all-bugs): report multiple violations per checkpoint - #180

Open
yeunmer2006 wants to merge 3 commits into
fmagent-project:mainfrom
yeunmer2006:feat/allbugs-v2
Open

feat(all-bugs): report multiple violations per checkpoint#180
yeunmer2006 wants to merge 3 commits into
fmagent-project:mainfrom
yeunmer2006:feat/allbugs-v2

Conversation

@yeunmer2006

Copy link
Copy Markdown
Contributor

Extend --all-bugs so that a single reasoning checkpoint can report multiple independently explainable violations, instead of returning at most one candidate per checkpoint.

Changes

  • Add an --all-bugs-specific structured response schema using a violations array.
  • Validate and normalize every returned violation.
  • Collect all violations reported by each reasoning checkpoint.
  • Preserve the existing single-verdict behavior when --all-bugs is disabled.
  • Update the CLI help text and English/Chinese documentation.

@yeunmer2006 yeunmer2006 closed this Aug 3, 2026
@yeunmer2006 yeunmer2006 reopened this Aug 3, 2026
Detect all-bugs reasoning artifacts independently of target count so zero-candidate MATCH workspaces do not fall back to stale validation results or retry traces. Continue deriving all-bugs totals solely from current bound candidate targets.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 8563639b4f

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread dashboard.py Outdated
Discover all-bugs candidate targets from workspace artifacts instead of the dashboard process's BUG_VALIDATION_MAX_RETRIES value. This keeps live and archived counters stable across shell configurations while preserving default-mode fallback.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant