# react-native-vision-camera-realtime

react-native-vision-camera-realtime guides low-latency VisionCamera v5 architecture for real-time video processing. It helps implement zero-copy GPU paths, Nitro plugins, and Skia or WebGPU overlays.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Windsurf (Devin Desktop), Claude.ai · 147 installs · ★ 181 repo stars
- Author: margelo
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/margelo/react-native-skills/tree/HEAD/skills/react-native-vision-camera-realtime/SKILL.md
- Updated: 2026-08-21
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/react-native-vision-camera-realtime

## What it does

The skill outlines constraints for building fast frame processing pipelines with VisionCamera v5. Its core principle keeps the entire data stream inside a single GPU context when drawing visual effects. Passing frame textures straight to `<SkiaCamera />` or WebGPU compute prevents expensive transfers back to system RAM. For native interop, it details integration patterns with Nitro HybridObjects and device accelerators.

Physical buffer rotation wastes rendering budgets. Setting `enablePhysicalBufferRotation: false` is required, passing `frame.orientation` and `frame.isMirrored` as metadata directly into matrix transforms. Calling methods like `getPixelBuffer()` triggers hard synchronization points across buses. This stalls the pipeline and leads to dropped frames.

When tasks must run asynchronously, in-flight work is strictly bounded to prevent queue growth. During development, developers can iterate quickly inside a Mac Catalyst target paired with an external USB camera before profiling hardware release builds.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add margelo/react-native-skills --skill react-native-vision-camera-realtime -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/react-native-vision-camera-realtime/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/react-native-vision-camera-realtime/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add margelo/react-native-skills --skill react-native-vision-camera-realtime -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/react-native-vision-camera-realtime/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add margelo/react-native-skills --skill react-native-vision-camera-realtime -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/react-native-vision-camera-realtime/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install margelo/react-native-skills skills/react-native-vision-camera-realtime
```

2. **Manually: put the files in place**
- File: `.github/skills/react-native-vision-camera-realtime/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add margelo/react-native-skills --skill react-native-vision-camera-realtime -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/margelo/react-native-skills --path skills/react-native-vision-camera-realtime
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/react-native-vision-camera-realtime/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add margelo/react-native-skills --skill react-native-vision-camera-realtime -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/margelo/react-native-skills/tree/HEAD/skills/react-native-vision-camera-realtime/SKILL.md

## When to use it
- prevent dropped frames during real-time video overlay rendering
- build zero-copy texture pipelines with VisionCamera v5 and WebGPU or Skia
- route camera frames into custom Nitro native plugins
- profile end-to-end frame latency across GPU and NPU backends

## Example prompts
- Review my VisionCamera v5 frame pipeline for unexpected CPU readbacks
- Configure a WebGPU texture pipeline that avoids physical buffer rotation
- Set up a bounded async runner pattern for ML inference on incoming camera frames

## Similar
- [react-native-skia](https://affpapa.org/en/nevibecode/skills/react-native-skia.md) — skill · 133 installs
- [react-native-best-practices](https://affpapa.org/en/nevibecode/skills/react-native-best-practices.md) — skill · 32700 installs
- [react-native-tv-best-practices](https://affpapa.org/en/nevibecode/skills/react-native-tv-best-practices.md) — skill · 2800 installs
- [expo-react-native-performance](https://affpapa.org/en/nevibecode/skills/expo-react-native-performance.md) — skill · 1500 installs
- [react-native](https://affpapa.org/en/nevibecode/skills/jezweb-react-native.md) — skill · 1300 installs
- [react-native-ease-refactor](https://affpapa.org/en/nevibecode/skills/react-native-ease-refactor.md) — skill · 958 installs

From the same set: [margelo/react-native-skills](https://affpapa.org/en/nevibecode/by/margelo/react-native-skills.md) — 9 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
