recursive-review

recursive-review is a command for Claude Code from fletchgqc/agentbox. It costs 11 tokens per session (326 once invoked), scanned A, original, Apache-2.0.

A command that runs several code-review passes, checks the first review, and combines the findings into a recommended path forward.

In plain words
What is it for?
Use it to review a requested code change through multiple stages and save the review, evaluation, and conclusion to files.
Why use it?
It reduces the chance of relying on one unchecked review or treating every suggestion as valid.

Command for Claude Code

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/fletchgqc/agentbox/recursive-review
Clone the repo
git clone --depth 1 https://github.com/fletchgqc/agentbox

Made for: Claude Code.

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 recursive-review

README.md
[![agentmods](https://agentmods.dev/badge/commands/fletchgqc/agentbox/recursive-review.svg)](https://agentmods.dev/commands/fletchgqc/agentbox/recursive-review)
Your own site
<a href="https://agentmods.dev/commands/fletchgqc/agentbox/recursive-review"><img src="https://agentmods.dev/badge/commands/fletchgqc/agentbox/recursive-review.svg" alt="Measured on agentmods" height="20"></a>
Per session 11 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 326 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 $0.00011 $0.00326
Opus 5 $0.00005 $0.00163
Sonnet 5 $0.00002 $0.00065
Haiku 4.5 $0.00001 $0.00033

Measured 4d ago against content hash 49bc13ff2b4a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

recursive-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 4d 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.

.claude/commands/recursive-review.md · 23 lines

What it actually says

You are a coordinator agent that orchestrates multiple specialized agents to conduct code review.

Steps:

Sequentially execute these three workflows:

  1. Launch a Task agent with the prompt: 'Follow the instructions in .claude/commands/review.md with argument "$1" and save the results to the file review/review1.md. It should be just one review for all mentioned changes. Report only the file name back to the main agent.'

  2. Launch a Task agent with the prompt: 'Conduct a critical evaluation of a code review. The review was performed on "$1" and is found in the file review/review1.md. Are the recommendations valid considering the project goals? Justify your reasoning. Save your results to review/evaluation1.md. It should be just one review for all mentioned changes. Report only the file name back to the main agent.'

  3. Launch a Task agent with the prompt: 'Conclude the best way forward for the change request described by "$1". A review was made and stored at review/review1.md. An evaluation of this review was stored at review/evaluation1.md. Objectively consider the recommendations made by each party and decide the way forward, explaining your reasoning. Document your recommendation in conclusion1.md. Report only the file name back to the main agent.'

  4. After all agents have completed, report summary of all processing

Important: Process files one at a time (serially) to avoid git branch conflicts. Do NOT launch agents in parallel.

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. 4d ago First seen · 23 lines · 11 tokens per session scan A 49bc13ff2b4a

Subscribe to this mod's changes

recursive-review is a command published in the GitHub repository fletchgqc/agentbox (113 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 11 tokens to every session and 326 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.