review

review is a command for Claude Code from 2mawi2/para. It costs 0 tokens per session (875 once invoked), scanned A, original, MIT.

Your task is to orchestrate the review and merge of the following para branches: $ARGUMENTS.

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/2mawi2/para/review
Clone the repo
git clone --depth 1 https://github.com/2mawi2/para

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 review

README.md
[![agentmods](https://agentmods.dev/badge/commands/2mawi2/para/review.svg)](https://agentmods.dev/commands/2mawi2/para/review)
Your own site
<a href="https://agentmods.dev/commands/2mawi2/para/review"><img src="https://agentmods.dev/badge/commands/2mawi2/para/review.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 875 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 $0.00000 $0.00875
Opus 5 $0.00000 $0.00438
Sonnet 5 $0.00000 $0.00175
Haiku 4.5 $0.00000 $0.00088

Measured 2d ago against content hash 635f114c040b, 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.

.claude/commands/review.md · 67 lines

What it actually says

Your task is to orchestrate the review and merge of the following para branches: $ARGUMENTS.

Phase 1: Parallel Review Use your internal Task tool to review each branch in parallel. IMPORTANT you must start all the initial review Tasks with your task tool in parallel for each review that can be performed in parallel. For each branch, start a new sub-task with the following instructions:

Sub-Task Prompt Template: "You are a code reviewer. Review the changes on the branch {{branch_name}} against the current branch (run 'git status' first to confirm current branch). CRITICAL: Do NOT check out the branch; only use git diff <current_branch>...{{branch_name}} (note the THREE dots - this shows only the changes made on the branch since it diverged, avoiding confusion from files added to main after the branch was created).

Provide a structured review response in the following format:

STATUS: OK or NEEDS_FIX

SUMMARY: (2-3 sentences describing what was implemented/changed)

COMPLETENESS: (List what was successfully completed)

  • Feature/change 1
  • Feature/change 2
  • etc.

ISSUES: (Only if NEEDS_FIX - list specific problems)

  • Issue 1: Description
  • Issue 2: Description
  • etc.

QUALITY: (Brief assessment of code quality, test coverage, and adherence to project standards)"

Phase 2: Follow-up Fixes (if needed)

  1. Wait for all reviews to complete.
  2. Provide Review Summary to User:
    • List all branches reviewed with their STATUS
    • For OK branches: Show the SUMMARY and COMPLETENESS sections
    • For NEEDS_FIX branches: Show SUMMARY, ISSUES, and planned fixes
    • Give user a brief overview before proceeding
  3. If any review returned NEEDS_FIX: a. Create follow-up task files with detailed fix instructions for each failed review b. Resume ALL failed sessions using para resume <session-name> --file <follow-up-task-file> --dangerously-skip-permissions c. Monitor progress: para list to see active sessions d. Wait for ALL agents to complete their fixes and finish their sessions

Phase 3: Final Review and Merge

  1. Once all follow-up fixes are complete (no more active para sessions): a. Re-review all previously failed branches to ensure issues are resolved b. If any still have issues, repeat Phase 2
  2. When all branches pass review: a. Merge branches sequentially into <current_branch>, resolving any conflicts b. Start with independent branches, then dependent ones c. After each successful merge, delete the local and remote feature branch

Follow-up Task Requirements When creating follow-up tasks for failed reviews, ensure each task includes:

  • Detailed explanation of what needs to be fixed based on the review feedback
  • Complete code examples for the fixes required
  • CRITICAL: Explicit instruction to commit all changes to the branch before calling para finish

Follow-up Task Template:

At the end of your follow-up task, always include: "After implementing all fixes:

  1. Commit all changes: git add . && git commit -m 'Fix [description of fixes]'
  2. Verify build works: just build
  3. Run: para finish '[commit message]'"

IMPORTANT:

  • Ensure to start each review TASK with our internal TASK in parallel if there are multiple reviews
  • Ensure that each TASK is instructed not to checkout the branch
  • Ensure that each TASK also knows the location of the original (or follow up) task file that the para dispatch agent was launched with if you know the task
  • For failed reviews, always resume sessions with follow-up tasks that include commit and finish instructions
  • Succeeded sessions that are not dependent on non succeeded sessions, can and should be merged directly and already integreated after the failed sessions where resumed.
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 · 67 lines · 0 tokens per session scan A 635f114c040b

Subscribe to this mod's changes

review is a command published in the GitHub repository 2mawi2/para (19 stars, last pushed 11mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 875 tokens. 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 2026-09-01.