Codex review

Codex Review

Independent gpt-5.6 review of local changes through the Codex CLI. Read-only - never lets Codex modify the repo. Works identically from the main loop, subagents, and workflow stages (no plugin runtime required).

Reviews run on gpt-5.6-sol (MODELS.md: code review = Sol) - pin BOTH the model (-m gpt-5.6-sol) and the effort (-c model_reasoning_effort="high", the v4 standard tier; "medium" only for trivial bounded diffs) on every command below rather than inheriting CLI defaults. For codex review, -m is a GLOBAL flag and must come before the subcommand. Paths written as <skill-dir> mean this skill's base directory (provided when the skill is invoked).

Mode selection

• Normal review (default): overall code-quality pass on a diff, same quality as running /review inside Codex.
• Adversarial review: when the user (or you) wants a specific decision, risk area, or design pressure-tested. Steerable with focus text; returns structured JSON.

Normal review

codex -m gpt-5.6-sol -c model_reasoning_effort="high" review --uncommitted   # staged + unstaged + untracked
codex -m gpt-5.6-sol -c model_reasoning_effort="high" review --base main     # branch vs base
codex -m gpt-5.6-sol -c model_reasoning_effort="high" review --commit <sha>  # a single commit

• Multi-file/large diffs take minutes: run via Bash run_in_background: true; the harness notifies on completion (no polling needed).
• Run as ONE bare command - no pipes, no cd && chains (keeps the RTK hook inert).

Adversarial review

1. Build the prompt from references/adversarial-prompt.md: fill {{TARGET_LABEL}} (e.g. "diff vs main"), {{USER_FOCUS}} (focus text or "general adversarial pass"), {{REVIEW_COLLECTION_GUIDANCE}} (usually: "Collect the diff yourself with git; read changed files fully."), and {{REVIEW_INPUT}} (either the diff inline, or an instruction to run git diff <base>...HEAD). Write it to the scratchpad.
2. Run as a single bare command, feeding the prompt file via stdin (- reads the prompt from stdin; a plain < file redirect keeps the command pipe-free):

codex exec -m gpt-5.6-sol -c model_reasoning_effort="high" -s read-only --cd <repo> --output-schema <skill-dir>/references/review-output.schema.json --output-last-message <scratchpad>/codex-adversarial-<ts>.json - < <scratchpad>/prompt.md

3. Read the output JSON: verdict (approve | needs-attention), findings[] (severity, title, body, file, line_start/line_end, confidence 0–1, recommendation), next_steps[].

Presenting results

• Present findings ranked by severity, verbatim in substance - do not soften.
• Never auto-apply fixes from a review. Present first; fix only when the user asks or when you already have an approved implementation task that covers it.
• Note the session can be reopened in Codex with codex resume if deeper follow-up is wanted.
• If Codex disagrees with a deliberate design choice, report the disagreement and your own position - don't silently adopt either side.

Effort

Always pin effort explicitly - sol's vendor default is low. Per MODELS.md (2026-07-24 recalibration): "high" (Sol worker) is the standard review tier; "medium" only for trivial bounded diffs. Never xhigh or ultra. If a Sol-high review still leaves material doubt, escalate per MODELS.md: Root resolves the hard part inline, or one independent gpt-5.6-terra "max" adversarial pass.

References adapted from openai/codex-plugin-cc (Apache-2.0; see references/NOTICE and the full license text in references/LICENSE).