review

A review checkpoint that asks Codex CLI to review your work when you try to exit. You can provide a focus and limit the number of review cycles.

In plain words
What is it for?
Use it to review security, error handling, or another chosen area of a code change before exiting.
Why use it?
It catches issues before the coding session ends and prevents an unfinished review from being overlooked.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/alleneubank/claude-code/review
Clone the repo
git clone --depth 1 https://github.com/alleneubank/claude-code
Per session 14 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,548 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00014 $0.01548
Opus 5 $0.00007 $0.00774
Sonnet 5 $0.00003 $0.00310
Haiku 4.5 $0.00001 $0.00155

Measured 2d ago against content hash b5c77dfe8b5e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/codex-reviewer/commands/review.md · 170 lines

How it starts

The opening of the file, as written. The whole thing — 170 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Start Codex Review Gate

Create a review gate that triggers Codex CLI review when you exit.

Parse Arguments

Arguments: $ARGUMENTS

Parse the following from arguments:

  • FOCUS: Everything before the first -- flag (the review focus)
  • --max-cycles: Number (default: 30) - maximum review cycles before auto-approve

Parsing rules:

  1. Text before any -- flags is the review focus
  2. Extract --max-cycles N if present (N must be a positive integer)
  3. If no focus text provided, use the default focus

Examples:

  • /codex-reviewer:review → default focus, max 30 cycles
  • /codex-reviewer:review "focus on security vulnerabilities" → security review, max 30 cycles
  • /codex-reviewer:review --max-cycles 3 → default focus, max 3 cycles
  • /codex-reviewer:review "verify error handling" --max-cycles 10 → error handling review, max 10 cycles

Step 0: Check for Existing Gate

IMPORTANT: Before creating a new gate, check if an ACTIVE one already exists. The stop hook checks the current repo AND walks up through all parent superprojects, looking for the first ACTIVE gate. An inactive file does not block - continue checking parents:

# Check current repo and all ancestor superprojects for ACTIVE gate
DIR="$(git rev-parse --show-toplevel 2>/dev/null || pwd)"
FOUND_ACTIVE=""
while [ -n "$DIR" ]; do
  STATE_FILE="$DIR/.claude/codex-review.local.md"
  if [ -f "$STATE_FILE" ] && grep -q "active: true" "$STATE_FILE"; then
    echo "Found ACTIVE state file at: $STATE_FILE"
    cat "$STATE_FILE"
    FOUND_ACTIVE="true"
    break
  fi
  # Check for parent superproject
  PARENT="$(git -C "$DIR" rev-parse --show-superproject-working-tree 2>/dev/null)"
  [ -z "$PARENT" ] && break
  DIR="$PARENT"
done
[ -z "$FOUND_ACTIVE" ] && echo "NO_ACTIVE_STATE_FILE"

If an ACTIVE state file exists (in current repo OR any ancestor):

The review gate is already active.

ALLOWED when gate is active:

  • Regenerate the handoff file (/handoff) to capture your latest work
  • The handoff file at ~/.handoffs/ can be updated freely

Read the full file on GitHub · 170 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 170 lines · 0 tokens per session scan A b5c77dfe8b5e

Subscribe to this mod's changes

review is a command published in the GitHub repository alleneubank/claude-code (52 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 14 tokens to every session and 1,548 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.