What is rn-device-qa?

A read-only process for gathering runtime evidence from a React Native or Expo app on a simulator, emulator, or physical device. Runtime evidence can include screenshots, logs, UI interactions, and exploratory quality checks.

Use it to plan acceptance checks, verify mobile screens and flows, or record reproducible quality findings.

It separates what can be proven from a running app from what can be inferred by reading source code, while avoiding changes to the device or external systems.

rn-device-qa — Skill for Claude Code by AndrewDongminYoo/rn-agents-kit

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.

Any agent

npx skills add AndrewDongminYoo/rn-agents-kit --skill rn-device-qa

Clone the repo

git clone --depth 1 https://github.com/AndrewDongminYoo/rn-agents-kit

Made for: Claude Code.

Or install rn-agents-kit, the plugin that ships this one along with the rest of its 10 skills.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/andrewdongminyoo/rn-agents-kit/rn-device-qa/github.svg)](https://agentmods.dev/skills/andrewdongminyoo/rn-agents-kit/rn-device-qa)

Your own site

<a href="https://agentmods.dev/skills/andrewdongminyoo/rn-agents-kit/rn-device-qa"><img src="https://agentmods.dev/badge/skills/andrewdongminyoo/rn-agents-kit/rn-device-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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/andrewdongminyoo/rn-agents-kit/rn-device-qa"><img src="https://agentmods.dev/badge/skills/andrewdongminyoo/rn-agents-kit/rn-device-qa.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 38 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,715 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00038	$0.01715
Opus 5.5	$0.00015	$0.00686
Sonnet 5.5	$0.00008	$0.00343
Haiku 4.5	$0.00004	$0.00171

Measured yesterday against content hash 3635e784ee35, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

rn-device-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 yesterday.

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.

skills/rn-device-qa/SKILL.md · 123 lines

How it starts

The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.

RN Device QA

Overview

This skill plans the smallest runtime evidence surface that can prove the requested acceptance criterion. It keeps static screenshots, JavaScript logs, interactive automation, and external state changes distinct. It does not operate a device, install tools, launch apps, or change external state.

When to Use

"verify this React Native screen on a device"
"run acceptance checks on the simulator or emulator"
"collect runtime evidence for this mobile flow"
"explore this app and record reproducible QA findings"

Prerequisites

Identify the requested app build and a read-only fingerprint source. Record the target platform, device or simulator identity, app identifier, account class, and starting state. Report build attribution as [UNCERTAIN] when no fingerprint source exists. Use an Android serial, iOS UDID, or one booted simulator identity when the evidence helper supports it.
For simulator capture, require exactly one booted simulator overall. Confirm that the booted simulator matches the recorded identity.
A Metro target descriptor carries no hardware identifier: it reports a title, a device name, and an app identifier only. Screenshot evidence is pinned by serial or UDID, so combining the two surfaces requires a read-only mapping from the recorded serial or UDID to the selected Metro target. Establish that mapping by confirming exactly one connected device or booted simulator reports the recorded device name, using adb devices -l for Android or xcrun simctl list devices booted for a simulator. Report device attribution as [UNCERTAIN] when two targets share a device name and app identifier, and do not present the screenshot and the log as evidence about the same device.
Record METRO_HOST as the selected Metro host:port before inspecting targets or collecting logs.
Record APP_ID as the expected app identifier for Metro preflight and collection descriptor verification.
Record METRO_SELECTOR as a unique device name or other unique target selector for Metro preflight and collection.
Set METRO_SKILL_DIR to the absolute base directory of the active rn-metro-console skill.
Confirm the requested evidence can be produced without changing external state, or identify the required separate authorization.
Identify a disposable fixture before a flow creates, edits, or deletes records.

Read the full file on GitHub · 123 lines

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.

yesterday Changed 3635e784ee35
2mo ago First seen · 123 lines · 38 tokens per session scan A 89d801f554c7

Subscribe to this mod's changes

rn-device-qa is a skill published in the GitHub repository AndrewDongminYoo/rn-agents-kit (3 stars, last pushed yesterday), licensed Apache-2.0. It adds 38 tokens to every session and 1,715 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-05.

Claude Code Mobile Testing agent-skills ai-coding claude-code-plugin code audit react-native

rn-device-qa next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
rn-device-qaAndrewDongminYoo/rn-agents-kit	Skill	3	yesterday	Apache-2.0	A	38	Claude Code
expoericrisco/rsc-harness	Skill	503	2d ago	MIT	A	79	Claude Code, Codex
react-nativeericrisco/rsc-harness	Skill	503	2d ago	MIT	A	90	Claude Code, Codex
init-projectwataru-maeda/react-native-boilerplate	Skill	524	5d ago	MIT	A	50	Claude Code, Codex
validate-mobilejohnkozaris/jko-claude-plugins	Skill	11	yesterday	MIT	A	20	Claude Code
ccc-mobileKevinZai/commander	Skill	6	yesterday	MIT	A	39	Claude Code
swiftui-expertjohnkozaris/jko-claude-plugins	Skill	11	yesterday	MIT	A	133	Claude Code