# document-intelligence

The document-intelligence skill routes document work through Snowflake Cortex AI, from field extraction and OCR to layout-aware parsing. It fits PDFs, scans, images, and office files.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Windsurf (Devin Desktop), Claude.ai · 1 install · ★ 1 repo stars
- Author: sfc-gh-nejain
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/sfc-gh-nejain/document-intelligence-skill/tree/HEAD/SKILL.md
- Updated: 2026-02-03
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/document-intelligence

## What it does

document-intelligence provides a workflow for document processing with Snowflake Cortex AI. It helps an agent identify the file type, clarify the desired output, and select structured extraction or full content parsing. Local files are uploaded to a Snowflake stage before processing. Results can then feed a table, pipeline, or RAG search.

The package points to `reference/ai-extract.md` for fields and tables and `reference/ai-parse-doc.md` for text, page structure, and images. Its process includes testing one file with `AI_EXTRACT`, reviewing the result, refining the request when needed, and moving to a batch run or storage step.

Supported inputs include PDF, PNG, JPEG, TIFF, DOCX, PPTX, HTML, and TXT. `AI_EXTRACT` also handles CSV, MD, and EML. XLSX and XLS cannot be passed directly to these document AI functions, while DOC and PPT need conversion first.

## Requirements
- Works with: Snowflake

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add sfc-gh-nejain/document-intelligence-skill --skill document-intelligence -a claude-code
```

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

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add sfc-gh-nejain/document-intelligence-skill --skill document-intelligence -a codex
```

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

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add sfc-gh-nejain/document-intelligence-skill --skill document-intelligence -a cursor
```

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

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install sfc-gh-nejain/document-intelligence-skill document-intelligence
```

2. **Manually: put the files in place**
- File: `.github/skills/document-intelligence/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add sfc-gh-nejain/document-intelligence-skill --skill document-intelligence -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/sfc-gh-nejain/document-intelligence-skill
```

### Windsurf (Devin Desktop)

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

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add sfc-gh-nejain/document-intelligence-skill --skill document-intelligence -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/sfc-gh-nejain/document-intelligence-skill/tree/HEAD/SKILL.md

## When to use it
- extract fields and tables from PDFs
- read text from scanned images
- parse invoices, contracts, and reports
- test extraction on one file before a batch run
- prepare document data for a table or RAG search

## Example prompts
- Use Snowflake `AI_EXTRACT` to pull fields and tables from this PDF.
- Parse this document with `AI_PARSE_DOCUMENT`, keeping its text and page structure.
- Run a test on one file, show the result, and refine the request before processing the batch.

## Similar
- [Amalgix Document Intelligence](https://affpapa.org/en/nevibecode/mcp/amalgix-document-intelligence.md) — MCP server · 3322 installs
- [azure-document-intelligence](https://affpapa.org/en/nevibecode/skills/azure-document-intelligence.md) — skill · 223 installs
- [azure-ai-document-intelligence-ts](https://affpapa.org/en/nevibecode/skills/azure-ai-document-intelligence-ts.md) — skill · 1 install
- [azure-ai-document-intelligence-dotnet](https://affpapa.org/en/nevibecode/skills/azure-ai-document-intelligence-dotnet.md) — skill · 1 install
- [platform-salesforce-connect-adapter-generate](https://affpapa.org/en/nevibecode/skills/platform-salesforce-connect-adapter-generate.md) — skill · 417 installs
- [cloud-finops](https://affpapa.org/en/nevibecode/skills/cloud-finops.md) — skill · ★ 75 repo stars

---
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
