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.
git clone --depth 1 https://github.com/sergeyklay/.agentsWrote 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/sergeyklay/.agents/challenge-pr)<a href="https://agentmods.dev/commands/sergeyklay/.agents/challenge-pr"><img src="https://agentmods.dev/badge/commands/sergeyklay/.agents/challenge-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 | $0.00012 | $0.00466 |
| Opus 5 | $0.00006 | $0.00233 |
| Sonnet 5 | $0.00002 | $0.00093 |
| Haiku 4.5 | $0.00001 | $0.00047 |
Grade A, and why
challenge-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 — 33 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Your task is to obtain an independent review of a pull request from a second model and arbitrate where that review and your own diverge.
Task
- Use the
challenge-prAgent Skill to drive every step: resolve the target, capture the diff, launch the second opinion, review independently, arbitrate, report. - Resolve the target from the argument: a PR number, URL, or
owner/repo#123. With no argument, use the current branch's open PR. - Read the project's context files (
AGENTS.md,CLAUDE.md,CONTRIBUTING.md,README.md) and architecture documentation, so your own pass judges the diff against this project's standards rather than generic ones.
Skill Enforcement
MANDATORY: Apply the challenge-pr Agent Skill verbatim.
The skill is the single source of truth for how the two reviews are obtained and reconciled. The project's context files are the source of truth for what counts as a defect worth reporting.
Process:
- Load the
challenge-prAgent Skill before reading any part of the diff. - Start the second opinion in the background before your own review begins, and do not open its output until your own findings are written down. The step order is what makes "Agreed" mean anything; reordering it produces a review that merely confirms another model.
- If the skill is unavailable in this environment, stop and report the failure. Do not improvise a replacement protocol and do not call any model provider directly.
- Emit the report in the chat response. The audience is the human operator, not a file.
Constraints
- Never modify code. No fixes, no commits, no pushes. The output is an opinion.
- Never post to GitHub. No review, no comment, no approval, no request-for-changes.
- Never report a finding from the second model without first checking it against the code yourself.
- Never treat the second model's severity or confidence as the verdict, and never let its unavailability stop the review.
- Name the model that actually served the request, taken from the script's output rather than from what was requested.
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 · 33 lines · 12 tokens per session scan A 0a9f5c522fa5
challenge-pr is a command published in the GitHub repository sergeyklay/.agents (5 stars, last pushed today), licensed Apache-2.0. It adds 12 tokens to every session and 466 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.
Other commands, from other repositories
proof-capture
Capture PR-ready proof artifacts for a feature, with an attested fail-closed controller in strict mode.
refactor-clean
You are a code refactoring expert specializing in clean code principles, SOLID design patterns, and modern software engineering best practices. Analyze and refactor the provided code to improve its quality, maintainability, and performance.
devkit.github.review-pr
Provides comprehensive GitHub pull request review with code quality, security, and best practices analysis. Use when reviewing a PR before merging.
sonarqube
Analyze SonarCloud quality issues for a specific PR.
refactor
Analyze code for SOLID violations and suggest targeted improvements.
test-feature
Test a React Native feature on the running simulator/emulator. Verifies UI, user flows, and internal state. Generates a persistent Maestro test file.