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.
npx skills add big-emotion/agent-atelier --skill review-traingit clone --depth 1 https://github.com/big-emotion/agent-atelierWrote 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/skills/big-emotion/agent-atelier/review-train)<a href="https://agentmods.dev/skills/big-emotion/agent-atelier/review-train"><img src="https://agentmods.dev/badge/skills/big-emotion/agent-atelier/review-train/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.
<a href="https://agentmods.dev/skills/big-emotion/agent-atelier/review-train"><img src="https://agentmods.dev/badge/skills/big-emotion/agent-atelier/review-train.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00179 | $0.03691 |
| Opus 5 | $0.00089 | $0.01845 |
| Sonnet 5 | $0.00036 | $0.00738 |
| Haiku 4.5 | $0.00018 | $0.00369 |
Grade A, and why
review-train 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 10d 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 — 207 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Review Train
Code-review a whole list of open pull requests in one pass — and only review. Enumerate every open PR, dispatch a team of parallel reviewer sub-agents, post one structured high-detail review comment per PR, and tag each PR with an approved or changes-requested label.
This skill is strictly review-scoped. It never fixes CI, never resolves merge conflicts, never pushes to any branch, never merges, never closes PRs. Its only writes are: one review comment per PR and one label per PR. It still follows a grouped-confirmation model — all analysis is done read-only first, one plan is presented, and it waits for a single explicit confirmation before posting anything.
If the user also wants CI fixes / conflict resolution / merging, that is a different job — point them at the merge-train skill instead. Do not blur the two.
When to Activate
- User shares a GitHub pull-requests URL and asks to review them (not merge).
- User says: "review all the open PRs", "code review the PR queue", "do a review pass on the open PRs", "review and label the PRs", "give me a detailed review on every open PR".
- User invokes
/review-train(optionally with a GitHub PRs URL or--dry-run).
If the user asks to merge / fix CI / resolve conflicts, this is the wrong skill — defer to merge-train.
Inputs
- Primary: a GitHub pull-requests URL. Any of:
/pulls, a search query (/pulls?q=...), a label/milestone filter, or a single PR URL. Honor the filter in the URL — only the PRs the URL would list are in scope. A single-PR URL ⇒ review just that one. - If no URL is given, default to the current repo's open PRs (
gh pr list); confirm the repo if ambiguous. - Flags:
--dry-run— produce the full plan + the drafted review bodies and stop; never post a comment or label.--reference <comment-url>— a GitHub comment URL whose structure and depth the reviews must match. If omitted, auto-detect (see Phase 0).--no-label— post review comments but do not apply/modify labels.--include-drafts— also review draft PRs (default: drafts are listed but skipped).--concurrency <n>— max parallel reviewer agents (default 4).--ci-timeout <minutes>— per-PR ceiling on how long to wait for CI to reach a definitive state before reviewing (default:30). Once the ceiling is hit, the PR is reviewed anyway withCI still pending after <timeout>recorded as context.--ci-poll <seconds>— interval betweengh pr checkspolls while waiting for CI (default:30).--no-wait-ci— skip the CI-settle wait entirely and review immediately on whatever CI state is current (may bepending). Default off.
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.
- 10d ago First seen · 207 lines · 179 tokens per session scan A 5728365f016a
review-train is a skill published in the GitHub repository big-emotion/agent-atelier (2 stars, last pushed 3d ago), licensed MIT. It adds 179 tokens to every session and 3,691 once invoked, about $0.0009 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.
Other skills, from other repositories
github-pr-review
Handles PR review comments and feedback resolution. Use when user wants to resolve PR comments, handle review feedback, fix review comments, address PR review, check review status, respond to reviewer, verify PR readiness, review PR comments, analyze review feedback, evaluate PR comments, assess review suggestions, or…
github-pr-creation
Creates GitHub Pull Requests with automated validation and task tracking. Use when user wants to create PR, open pull request, submit for review, or check if ready for PR. Analyzes commits, validates task completion, generates Conventional Commits title and description, suggests labels. NOTE - for merging existing…
ci-review
CI-optimized code review: multi-agent parallel review with confidence scoring and atomic GitHub PR review posting. Runs specialized review agents, scores findings, and submits a GitHub PR review via post-review.sh. Always executes post-review.sh to submit a review on every run (including zero-findings runs). Triggers…
babysit
PR babysitter: monitors CI status, auto-rebases when behind, auto-fixes CI where possible, delegates review comment handling to dlc:pr-check, and re-requests review after fixes. Designed for /loop usage with Remote Control.
web-security-review
Top-level security code review for JavaScript/TypeScript web applications and Electron desktop applications. Use when the user asks to "security review", "find vulnerabilities", "check for security issues", "audit security", "OWASP review", "is this secure?", or to review…
second-opinion
Get a second opinion on freshly written code from the other coding agent — Claude asks Codex, Codex asks Claude — then empirically vet every finding against ground truth before accepting or vetoing it. Use once the code is written and the tests pass, before you commit. Do not use for a delegated review or…