A collection of agent-optimized React Native skills for AI coding assistants, this project provides structured, actionable instructions for domain-specific tasks, primarily focusing on performance optimization. It targets developers using AI coding assistants, enabling them to receive precise, context-aware guidance on complex React Native development challenges, thereby improving efficiency and code quality.

How It Works

The project packages domain-specific knowledge, particularly React Native optimization best practices, into standardized "skills." These skills are designed to be discoverable and executable by various AI coding assistants. The core approach leverages a structured format (following the Claude Code plugin standard) that allows AI models to interpret and apply the provided instructions, offering concrete steps for performance tuning across JavaScript/React, native modules, and bundling.

Quick Start & Requirements

Installation varies by AI assistant:

• Claude: Use /plugin marketplace add callstackincubator/agent-skills and /plugin install react-native-best-practices@callstack-agent-skills.
• Cursor: Install from GitHub URL (https://github.com/callstackincubator/agent-skills.git) or clone locally.
• OpenAI Codex CLI: Use $skill-installer install react-native-best-practices from callstackincubator/agent-skills or clone manually.
• Gemini CLI: Use gemini skills install https://github.com/callstackincubator/agent-skills.git.
• OpenCode: Clone to .opencode/skill/ or ~/.config/opencode/skill/. No specific local development prerequisites are detailed beyond those required by the AI assistants themselves.

Highlighted Details

• Provides React Native optimization best practices derived from Callstack's "The Ultimate Guide to React Native Optimization."
• Covers optimization areas including JavaScript/React (profiling, re-renders), Native (iOS/Android profiling, TTI), and Bundling (analysis, app size).
• Designed for interoperability with major AI coding assistants like Cursor, GitHub Copilot, and Claude.

Maintenance & Community

The project is presented as an evolving initiative ("just the start") with an open invitation for contributions. A roadmap item includes planned integration for visual feedback processing. Contact is available via hello@callstack.com.

Licensing & Compatibility

The specific open-source license for this repository is not explicitly stated in the provided README. Compatibility is broad, targeting multiple AI coding assistants.

Limitations & Caveats

Key planned features, such as integrating Model Context Protocol (MCP) for visual profiler output analysis (e.g., flame graphs, treemaps), are not yet implemented. This means AI agents currently cannot autonomously interpret visual performance data. The project is in its early stages, indicating potential for ongoing changes and feature development.