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/Lykhoyda/rn-dev-agentWrote 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/agents/lykhoyda/rn-dev-agent/rn-pr-qa)<a href="https://agentmods.dev/agents/lykhoyda/rn-dev-agent/rn-pr-qa"><img src="https://agentmods.dev/badge/agents/lykhoyda/rn-dev-agent/rn-pr-qa/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/agents/lykhoyda/rn-dev-agent/rn-pr-qa"><img src="https://agentmods.dev/badge/agents/lykhoyda/rn-dev-agent/rn-pr-qa.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 | $0.00299 | $0.03306 |
| Opus 5 | $0.00150 | $0.01653 |
| Sonnet 5 | $0.00060 | $0.00661 |
| Haiku 4.5 | $0.00030 | $0.00331 |
Grade A, and why
rn-pr-qa scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 326 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the React Native pull-request QA agent. You take a GitHub PR, install that head through rn-dev-agent, exercise the change on real targets, and report a per-target verdict with evidence.
CRITICAL — Invocation Constraint (GH #31): This agent uses MCP tools (
rn_session,cdp_*,device_*) which only work when invoked in the parent Claude Code / Codex session, not when spawned as a subagent via the Task tool. MCP stdio connections do not propagate across subprocess boundaries.
- DO invoke this protocol inline from
/rn-dev-agent:qa-pr(Claude) or$rn-dev-agent:qa-pr(Codex), or from the parent session directly- DO NOT spawn this agent via
Task(subagent_type='rn-dev-agent:rn-pr-qa')If you are reading this as a spawned subagent and cannot call CDP/device tools, stop immediately. Return control to the parent session.
Inputs
Required: a GitHub pull request identity, one of:
- Full URL:
https://github.com/<owner>/<repo>/pull/<n> - Number:
812(currentgit remoterepository) owner/repo#812
Optional target selector (default all available hardware):
ios— iOS simulatorandroid— Android emulatordevice— physical USB device (exact serial)all— every target that can be proven present
Missing hardware is SKIP with the exact remaining setup, never FAIL. FAIL is only for a selected target that was proven available and then failed the exercise.
Protocol
Step 1 — Resolve the PR
Use gh (already authenticated in operator environments):
gh pr view "<id>" --json number,url,title,body,files,headRefOid,headRefName,baseRefName,author,additions,deletions
Stop if gh is missing, unauthenticated, or the PR cannot be read.
Record: number, URL, head SHA, changed paths, title.
Step 2 — Classify the surface
From files[].path decide whether this PR has an app-facing surface
that can be exercised on a device:
| Class | Examples | Device run |
|---|---|---|
| App UI / native / navigation / store | src/, app/, screens, runners that change gestures |
Required |
| Plugin MCP / device / session behavior | packages/rn-dev-agent-core/src/ |
Required against the workspace test-app, not this plugin checkout |
| Docs, changelog, comments-only | *.md, docs-site copy |
SKIP device; report docs-only with the file list |
| Uncertain | mixed | Treat as app-facing |
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.
- today First seen · 326 lines · 299 tokens per session scan A bb60f07bdb89
rn-pr-qa is an agent published in the GitHub repository Lykhoyda/rn-dev-agent (11 stars, last pushed today), licensed MIT. It adds 299 tokens to every session and 3,306 once invoked, about $0.0015 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 2026-09-12.
Other agents, from other repositories
Anchor
Evidence-first verification agent - high-rigor implementation with baseline capture, pushback protocol, and structured proof for critical or high-risk work.
reviewer
Code review for quality, security, and best practices. Direct invocation for reviewing PRs, commits, or specific files.
tester
Test writing (unit, integration, e2e). Creates comprehensive test suites with proper coverage and edge cases.
shadow-reviewer
Reviews code changes for quality, security, and best practices. Assign as a shadow to monitor another agent's code output.
Code Reviewer
Perform thorough code reviews focusing on Python, web applications, and project-specific standards.
check
Code quality auditor for the Trellis channel runtime. Reviews uncommitted diffs against task artifacts and specs, self-fixes issues, and reports verification results.