From review-by-opp
Addresses open Codex review findings from reviews/current.json by fixing code changes or marking explicit resolutions (fixed, wont_fix, duplicate, etc.). Reports progress.
How this skill is triggered — by the user, by Claude, or both
Slash command
/review-by-opp:fixThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You are addressing Codex review findings. You are the builder - fix what needs fixing.
You are addressing Codex review findings. You are the builder - fix what needs fixing.
Load the ledger:
reviews/current.json/review-by-opp:start first."Show open findings:
open, grouped by severity (critical first)Address findings one by one: For each finding, you MUST do one of these - no exceptions:
a. Fix it → Make the code change, then update the finding:
status to fixedresolution_note to a brief description of what you changedb. Mark as won't_fix → If the finding is intentional or acceptable:
status to wont_fixresolution_note explaining whyc. Mark as not_reproducible → If you cannot reproduce the issue:
status to not_reproducibleresolution_note with your investigationd. Mark as needs_context → If you need more information:
status to needs_contextresolution_note describing what context is needede. Mark as duplicate → If this duplicates another finding:
status to duplicateduplicate_of to the other finding's IDUpdate the ledger:
reviews/current.jsonReport progress:
/review-by-opp:fix again or /review-by-opp:review for another Codex pass"/review-by-opp:review) or finalize (/review-by-opp:finalize)"CRITICAL: You MUST NOT claim "all fixed" unless every finding has an explicit resolution state. The ledger enforces this - hand-waving is not allowed.
$ARGUMENTS
npx claudepluginhub akshan-main/review-by-opp --plugin review-by-oppFixes all open review findings in a single pass by fetching job IDs from the roborev CLI and applying suggested fixes to the codebase.
Handles code review feedback rigorously: verifies suggestions, resolves unclear items, implements fixes in priority order. Useful when addressing review comments.
Iterative code review loop that runs multiple rounds of review and auto-fixes, using /deep-review for safe inline fixes and /big-plan for larger changes. Deferred findings can become GitHub issues.