codex-review

name

codex-review

description

Ask Codex CLI (gpt-5.5) for an independent code review of uncommitted changes, a branch diff, a commit, or a specific implementation. Use when the user asks for Codex or gpt-5.5 review work, when the model-selection rubric calls for a gpt-5.5 review perspective, or when Codex should audit a diff, find bugs or regressions, or compare an implementation against requirements. For a review by the current agent itself, use the normal review process instead.

Codex Review

Use Codex as an independent second-opinion reviewer. Per the model rubric, primary reviews belong to fable or opus (/code-review, review-swarm); this skill adds the gpt-5.5 perspective on top of those, or handles the case where the user explicitly asks for a Codex review.

Prefer the current agent's normal review process for small local checks. Do not delegate every lint, type, or formatting issue to Codex.

Workflow

1. Identify the review target: uncommitted changes, base branch, commit SHA, PR checkout, or specific files.
2. Create a temporary artifact directory for the Codex report and log.
3. Run Codex review with a focused review prompt.
4. Read Codex's report and verify important claims against the code before presenting them.

codex exec review has no -C flag — run it from the repository being reviewed. Reviews are read-only and quality matters, so run them at xhigh effort.

ARTIFACT_DIR="$(mktemp -d "${TMPDIR:-/tmp}/codex-review.XXXXXX")"
REPORT="$ARTIFACT_DIR/report.md"
PROMPT="$ARTIFACT_DIR/prompt.md"
LOG="$ARTIFACT_DIR/codex.log"

Review staged, unstaged, and untracked changes:

codex exec review -c model_reasoning_effort=xhigh --uncommitted \
  -o "$REPORT" - < "$PROMPT" > "$LOG" 2>&1

Review current branch against a base branch:

codex exec review -c model_reasoning_effort=xhigh --base <branch> \
  -o "$REPORT" - < "$PROMPT" > "$LOG" 2>&1

Review a single commit:

codex exec review -c model_reasoning_effort=xhigh --commit <sha> \
  -o "$REPORT" - < "$PROMPT" > "$LOG" 2>&1

xhigh reviews take a while — launch via Bash run_in_background; the findings land in $REPORT via -o.

Review Prompt

Ask Codex to use a code-review stance:

Review these changes for bugs, regressions, missing tests, security issues, and requirement mismatches.

Prioritize findings over summary. For each finding include:
- severity
- file and line reference
- concrete failure mode
- suggested fix direction

Do not edit files. If there are no substantive findings, say so and name any residual test gaps.

Add task-specific context when useful: requirements, risky areas, expected behavior, relevant tests, or files the current agent is unsure about.

Follow-ups

To ask Codex to elaborate on a finding or re-check after a fix, resume its session. Codex prints a session id: line in its output frontmatter — resuming by id keeps follow-ups pinned to this run even if other Codex sessions ran in between:

SESSION_ID="$(sed -n 's/^session id: //p' "$LOG" | head -1)"
codex exec resume "$SESSION_ID" -o "$REPORT" "<question or re-check request>"

Reporting Back

Before relaying a Codex finding, inspect the cited code or diff enough to decide whether the finding is real. In the user-facing response, separate confirmed issues from Codex suggestions you did not verify.

If Codex finds nothing, say that clearly and mention what review target it inspected.

If codex is not installed or the command fails, report the error and offer to review the changes directly instead.