React Native Agents gives your AI coding tool specialist React Native guidance instead of generic React advice. The agents inspect the project context first, use deep references on demand, and explain findings with practical fixes and verification steps.

npx @maheshwarimrinal/react-native-agents install

Or install for one tool:

npx @maheshwarimrinal/react-native-agents install --tool cursor
npx @maheshwarimrinal/react-native-agents install --tool claude-code
npx @maheshwarimrinal/react-native-agents install --tool windsurf

Then ask your assistant a normal React Native question:

The catalogue stutters when I scroll on Android. Find the likely causes and show me how to measure them.

The right specialist is selected automatically, or you can invoke one directly using the tool's supported command or chat mode.

The GitHub Action reviews changed files using the relevant specialists and posts findings to the pull request.

- name: React Native audit
  uses: maheshwarimrinal/react-native-agents@v1.0.3
  with:
    provider: anthropic
    model: claude-sonnet-5
    api-key: ${{ secrets.ANTHROPIC_API_KEY }}
    budget-usd: '1.50'
    fail-on: never

OpenAI is also supported:

- name: React Native audit
  uses: maheshwarimrinal/react-native-agents@v1.0.3
  with:
    provider: openai
    model: gpt-5
    api-key: ${{ secrets.OPENAI_API_KEY }}

Store API keys as GitHub Actions repository secrets. Do not commit them to workflow files.

A failed run fails the check. If an agent errors — an invalid key, no credits left, a provider outage — the audit exits non-zero rather than reporting a pass with zero findings. A green check that means "we did not actually look" is worse than a red one. Set fail-on-error: false if you would rather accept partial coverage.

Day to day — these respond to a question, an error log, or a request:

Review — these run on a diff, and on pull requests:

Doctor and Build are deliberately excluded from pull-request routing — they need something brought to them, so firing them at a diff would spend tokens to say nothing.

9 playbooks and 52 reference documents. Knowledge is verified through React Native 0.87 and Expo SDK 57; always verify version-specific advice against your project.

rn-size analyses your production bundle from its source map. It is deterministic — no model call, no API key, no cost, so it can run on every pull request for free.

npx @maheshwarimrinal/react-native-agents size
npx @maheshwarimrinal/react-native-agents size --base main --budget-delta 100kb

📦 Bundle size 📈 +332.0 KB
2.00 MB → 2.33 MB (+16.2%)

❌ Budget exceeded — increase 332.0 KB exceeds budget 100.0 KB

| moment 🆕 | +207.0 KB | 207.0 KB |
→ date-fns or dayjs (Hermes ships Intl, so formatting may need no library at all)

Bytes are attributed per source-map segment, so a package's number is measured rather than estimated. Budgets accept 250kb, 1.5mb, or a raw byte count; an invalid value is an error rather than a silently skipped check.

• Project page — overview of the agents, the bundle-size analyzer, and the design decisions behind them
• Real-world demo — intentionally flawed React Native project with a PR audit workflow
• Installation by tool — Claude Code, Cursor, Windsurf, Copilot, Codex, and MCP
• GitHub Action guide — providers, secrets, inputs, outputs, routing, and severity gates
• Agents guide — what each specialist catches and when to use it
• Usage guide — prompts, commands, MCP, full audits, severity, and finding format
• Architecture — source files, generator, generated targets, and design decisions
• Development guide — build, test, evals, freshness checks, and extending the project
• FAQ — versions, Expo, API keys, customization, npm, and troubleshooting
• Contributing
• Security policy

MIT — see LICENSE.

If React Native Agents saves you time, you can support its ongoing maintenance: