AReaL is an infrastructure system for training large language models with reinforcement learning, connecting model training to applications built around AI agents. Researchers and developers use it to train reasoning and agentic models through asynchronous workflows, and the catalogue add-ons support working with AReaL.
Borrowing it
Nothing to install: this file belongs to areal-project/AReaL. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/areal-project/AReaL/main/.claude/commands/review-pr.mdgit clone --depth 1 https://github.com/areal-project/AReaLWrote 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.
[](https://agentmods.dev/commands/areal-project/areal/review-pr)<a href="https://agentmods.dev/commands/areal-project/areal/review-pr"><img src="https://agentmods.dev/badge/commands/areal-project/areal/review-pr.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00017 | $0.02273 |
| Opus 5 | $0.00009 | $0.01137 |
| Sonnet 5 | $0.00003 | $0.00455 |
| Haiku 4.5 | $0.00002 | $0.00227 |
Grade A, and why
review-pr 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 8d 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.
How it starts
The opening of the file, as written. The whole thing — 294 lines — stays where its author put it; the contents beside it link to each section on GitHub.
@.claude/data/review-pr-domains-and-signals.md @.claude/data/review-pr-templates.md
PR Code Review (Dynamic Agent Allocation)
Intelligent code review for the current branch's Pull Request. Dynamically generates targeted review tasks based on PR changes.
Arguments
$ARGUMENTS
- No arguments: Review PR for current branch
- PR number: Review specific PR (e.g.,
/review-pr 123) --quick: Quick mode, only run Phase 1 analysis
Quick Start
- Get current branch PR:
gh pr view --json number,title,state,isDraft - If PR doesn't exist or is closed, stop and explain
- Execute Phases 1-4 in order
Workflow Overview
Phase 1: Deep PR Analysis [Haiku + Sonnet]
|- 1.0 PR Status Check [Haiku]
|- 1.1 Get PR Summary [Haiku]
+- 1.2-1.4 Domain/Signal Detection [Sonnet]
|
Phase 2: Dynamic Agent Planning [Sonnet]
|
Phase 3: Execute Review Tasks [Parallel, Dynamic Model Selection]
|
Phase 4: Confidence Scoring & Summary [Haiku]
Model Configuration
| Mode | CRITICAL/HIGH | MEDIUM | LOW |
|---|---|---|---|
| Default | Opus | Sonnet | Haiku |
Quick (--quick) |
Sonnet | Sonnet | Sonnet |
Economy (--economy) |
Sonnet | Haiku | Haiku |
Phase 1: Deep PR Analysis
1.0 PR Status Check [Haiku]
Check if PR should be reviewed:
- Is it closed? -> Stop
- Is it a draft? -> Note but continue
- Is it bot-generated? -> Skip
1.1 Get PR Summary [Haiku]
Get basic PR info: title, description, modified files, change summary.
1.2 Domain & Signal Detection [Sonnet]
Analyze each file change, detecting L1 domains and L2 signals by risk level.
Reference: See review-pr-domains-and-signals.md for complete domain tables:
- L1 domains (Distributed Runtime, Model Compute & Attention, Inference Backend & Serving, etc.)
- L2 signals per domain
- cross-domain linkage rules
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.
- 8d ago First seen · 294 lines · 17 tokens per session scan A 3b7b8f4c68fb
review-pr is a command published in the GitHub repository areal-project/AReaL (5,736 stars, last pushed today), licensed Apache-2.0. It adds 17 tokens to every session and 2,273 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.
Other commands, from other repositories
review-prs
Parallel PR triage: decide merge-worthiness, prepare rebased worktrees, fix blockers, return them for human merge.
audit
Audit a window of recent commits for real defects, confirm each finding before fixing, verify every surface a fix touches, then report confirmed/dismissed/no-op with evidence. Enforces Law 4 (Verify Before Reporting).
pr
Open a pull request with a description generated from the diff.
ship
Pre-ship pipeline — test, review, then prepare a clean commit.
pr-shepherd
Take every open PR end-to-end — diagnose, fix mechanical failures, stop at the merge gate.
gsd:inbox
Triage and review open GitHub issues and PRs against project templates and contribution guidelines.