# cdf-curation-and-retrieval

The cdf-curation-and-retrieval skill guides video and image curation with Cosmos Curator, plus handoffs to TAO and searches through Cosmos Dataset Search. Use it for requests involving CDF or its Make and CLI workflows.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf (Devin Desktop), Claude.ai · ★ 4106 repo stars
- Author: NVIDIA
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/NVIDIA/skills/tree/HEAD/skills/cdf-curation-and-retrieval/SKILL.md
- Updated: 2026-10-23
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/cdf-curation-and-retrieval

## What it does

The skill helps shape video collections by splitting clips, filtering material, removing duplicates, and preparing shards. For images, its workflow covers annotation and embeddings. Curator outputs can feed later data mining in TAO, while Cosmos Dataset Search has its own ingest and search route.

Before producing pipeline YAML, the agent checks for KPI results, sample data, or a calibration interview. An explicit run also requires config validation and user authorization. The setup calls for a GPU host with NVIDIA drivers and `nvidia-container-toolkit`, Docker, and access to the Makefile and the `configs/`, `cookbook/`, and `references/` trees. Without that GPU setup, a pipeline run cannot proceed. Workplace monitoring of employee conduct requires legal review.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add NVIDIA/skills --skill cdf-curation-and-retrieval -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/cdf-curation-and-retrieval/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/cdf-curation-and-retrieval/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add NVIDIA/skills --skill cdf-curation-and-retrieval -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/cdf-curation-and-retrieval/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add NVIDIA/skills --skill cdf-curation-and-retrieval -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/cdf-curation-and-retrieval/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install NVIDIA/skills skills/cdf-curation-and-retrieval
```

2. **Manually: put the files in place**
- File: `.github/skills/cdf-curation-and-retrieval/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add NVIDIA/skills --skill cdf-curation-and-retrieval -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/NVIDIA/skills --path skills/cdf-curation-and-retrieval
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/cdf-curation-and-retrieval/SKILL.md` — The legacy .windsurf/skills/ path still works

### 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/NVIDIA/skills/tree/HEAD/skills/cdf-curation-and-retrieval/SKILL.md

## When to use it
- split video into clips and remove duplicates
- filter video collections and divide them into shards
- prepare annotated images with embeddings
- validate Curator output before handing it to TAO
- ingest datasets into Cosmos Dataset Search and search them

## Example prompts
- Set up a Cosmos Curator video pipeline that splits clips and removes duplicates.
- Check what context is needed before writing YAML for an image pipeline.
- Validate the Curator output and prepare it for a TAO handoff.

## Similar
- [runpod](https://affpapa.org/en/nevibecode/skills/runpod.md) — skill · 1400 installs
- [dataset-curation](https://affpapa.org/en/nevibecode/skills/dataset-curation.md) — skill · 2600 installs
- [remove-ai-watermarks](https://affpapa.org/en/nevibecode/skills/remove-ai-watermarks.md) — skill · ★ 6481 repo stars
- [Remove AI Watermarks](https://affpapa.org/en/nevibecode/mcp/remove-ai-watermarks.md) — MCP server · ★ 6481 repo stars
- [mcp](https://affpapa.org/en/nevibecode/mcp/wavespeedai-mcp.md) — MCP server · ★ 38 repo stars
- [Runway ML](https://affpapa.org/en/nevibecode/mcp/runway-ml.md) — MCP server · ★ 14 repo stars

From the same set: [nvidia/skills](https://affpapa.org/en/nevibecode/by/nvidia/skills.md) — 312 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
