# RAG Document Q&A

> Use this tool when you need to generate accurate answers to questions based on private document collections. It solves the problem of retrieving relevant information from large PDF files, providing source-grounded and hallucination-free responses. The tool takes in document queries as input and outputs precise answers, making it ideal for applications requiring reliable document-based question-answering.

Canonical page: https://skillsregistry.net/skills/tanmayrawal-rag-document-qa  
JSON: https://api.skillsregistry.net/v1/skills/tanmayrawal-rag-document-qa

## Description

A RAG-powered MCP server for document question-answering. Indexes PDFs using FAISS vector search and Sentence Transformers, then generates grounded answers via Groq LLaMA inference. Includes a Streamlit chat interface and Claude Desktop integration for querying private document collections with source-grounded, hallucination-free responses.

## Trust

- **Trust score (0–1):** 0.88
- **Verification tier:** verified
- **Last scanned:** 2026-09-19

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** ai-ml
- **Updated:** 2026-09-19

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/tanmayrawal-rag-document-qa)
- **Repository:** <https://github.com/tanmayrawal/rag-mcp-server>

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "tanmayrawal-rag-document-qa"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/tanmayrawal-rag-document-qa` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/tanmayrawal-rag-document-qa/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt