review-dual

review-dual is a skill for Claude Code from anhnguyen0905/codex-mcp. It costs 24 tokens per session (1,543 once invoked), scanned A, original, MIT.

A review process that asks both Codex and Claude to inspect work independently, then compares their findings using evidence.

In plain words
What is it for?
Use it for code reviews that compare findings, resolve disagreements, and record improvements that do not block the task.
Why use it?
It reduces the chance that one reviewer misses a problem or approves its own changes without an independent check.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: mentions subagents; names the AskUserQuestion tool; mentions Codex.

Part of the codex-flow plugin — 65 skills, 2 commands, 1 MCP server shipped together

Good fit Use it for code reviews that compare findings, resolve disagreements, and record improvements that do not block the task.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/anhnguyen0905/codex-mcp/review-dual
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.

Any agent
npx skills add anhnguyen0905/codex-mcp --skill review-dual
Clone the repo
git clone --depth 1 https://github.com/anhnguyen0905/codex-mcp

Made for: Claude Code.

Or install codex-flow, the plugin that ships this one along with the rest of its 65 skills, 2 commands, 1 MCP server.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for review-dual

README.md
[![agentmods](https://agentmods.dev/badge/skills/anhnguyen0905/codex-mcp/review-dual/github.svg)](https://agentmods.dev/skills/anhnguyen0905/codex-mcp/review-dual)
Your own site
<a href="https://agentmods.dev/skills/anhnguyen0905/codex-mcp/review-dual"><img src="https://agentmods.dev/badge/skills/anhnguyen0905/codex-mcp/review-dual/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for review-dual

Your own site · 80×15
<a href="https://agentmods.dev/skills/anhnguyen0905/codex-mcp/review-dual"><img src="https://agentmods.dev/badge/skills/anhnguyen0905/codex-mcp/review-dual.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 24 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,543 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00024 $0.01543
Opus 5 $0.00012 $0.00772
Sonnet 5 $0.00005 $0.00309
Haiku 4.5 $0.00002 $0.00154

Measured yesterday against content hash a3a6abca2899, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

review-dual 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 yesterday.

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.

skills/review-dual/SKILL.md · 94 lines

How it starts

The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Dual Review (Codex + Claude)

Give every task TWO independent reviews: Claude reviews with the review-conformance, review-quality, and review-security skills; Codex reviews via mcp__codex__codex_review in a fresh read-only session so it never grades its own homework. Compare both reviews and select the optimal resolution supported by evidence.

Codex reviewer role (focus block)

Paste this template into the focus parameter of mcp__codex__codex_review. Without checkpoint commits, each per-task mcp__codex__codex_review sees the cumulative uncommitted diff, so the focus block MUST name and enforce the task's Files: scope:

Role: Act as an independent senior reviewer.
Task: <task id> — <task title>
Acceptance criteria: <paste the task's acceptance criteria>
Files: <paste the task's Files: list>
Review only the task's Files: list; report no findings outside it.
Review in this exact order:
1. Conformance to the task's .codex-flow/CONTEXT-T<n>.md slice and acceptance criteria.
Treat [verify]-stamped context as hypotheses to re-confirm against the current code before relying on it.
Read full .codex-flow/PLAN.md only as escalation when a finding disputes plan intent.
2. Quality: correctness hazards, silent failures, edge cases, and test quality.
3. Security: secrets, injection, validation, and unsafe patterns.
Output numbered findings tagged CRITICAL, HIGH, MEDIUM, or LOW.
For every finding, give file:line plus expected vs observed behavior.
Then add a SEPARATE `## Improvements` section.
Tag non-blocking suggestions IMP-1, IMP-2, ...: worthwhile refactors, better naming, missing nice-to-have tests, or documentation gaps that are not defects.

Pass the task's scope, and route out-of-scope findings to the ledger

Call mcp__codex__codex_review with scope: { files: <the task's Files: list>, contract: <the PLAN Contracts the task lists> } alongside focus. The server appends the contract and the file list to the reviewer prompt and stamps each returned finding with inScope, counting the rest in reviewFindings.outOfScopeCount.

Read the full file on GitHub · 94 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. yesterday Changed · +12 lines a3a6abca2899
  2. 2d ago Changed · +4 lines 7f9a96fa8e31
  3. 7d ago Changed · +20 lines a75af14e9f12
  4. 11d ago First seen · 58 lines · 24 tokens per session scan A 8df354e869b9

Subscribe to this mod's changes

review-dual is a skill published in the GitHub repository anhnguyen0905/codex-mcp (3 stars, last pushed yesterday), licensed MIT. It adds 24 tokens to every session and 1,543 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 2026-08-31.