rag-documents

name	rag-documents
description	Analyze uploaded documents (PDF, Word, Excel, PPT) using local RAG system at http://localhost:9000. Use when user uploads a file and asks to analyze, summarize, or extract information from it.

RAG Document Analysis

Analyze uploaded documents using the local RAG system.

When to Use

• User uploads a file and asks to analyze/summarize it
• User asks to extract data from Excel/PDF
• User asks questions about document content

Supported Formats

• Excel (.xlsx, .xls)
• PDF (.pdf)
• Word (.docx)
• PowerPoint (.pptx)

Prerequisites

1. Get API Token

First, get a token from the RAG system:

# Register a user (one-time)
curl -X POST "http://localhost:9000/v1/users/register" \
  -H "Content-Type: application/json" \
  -d '{"username": "testuser", "email": "[e-mail]", "password": "testpass123"}'

# Or login to get token
curl -X POST "http://localhost:9000/v1/users/login" \
  -H "Content-Type: application/json" \
  -d '{"username": "testuser", "password": "testpass123"}'

Save the access_token from response.

2. Create Project

curl -X POST "http://localhost:9000/v1/projects" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $TOKEN" \
  -d '{"name": "my-project", "description": "For document analysis"}'

Save the id from response as PROJECT_ID.

Usage Steps

1. Upload Document

curl -X POST "http://localhost:9000/v1/documents/upload?project_id=$PROJECT_ID" \
  -H "Authorization: Bearer $TOKEN" \
  -F "file=@/path/to/file.xlsx"

Returns: {"document_id": "xxx", "status": "processing"}

2. Wait for Processing

# Poll until status is "completed"
curl "http://localhost:9000/v1/documents/{document_id}"

Wait 2-3 seconds between polls. Status: processing → completed

3. Analyze with RAG

curl -X POST "http://localhost:9000/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen2.5-14b",
    "messages": [{"role": "user", "content": "Summarize this document"}],
    "project_id": "temp",
    "enable_rag": true,
    "top_k": 5
  }'

4. Cleanup (Optional)

curl -X DELETE "http://localhost:9000/v1/documents/{document_id}"

Complete Example

# Upload file
RESPONSE=$(curl -s -X POST "http://localhost:9000/v1/documents/upload?project_id=temp" \
  -F "file=@uploaded_file.xlsx")
DOC_ID=$(echo $RESPONSE | python3 -c "import sys,json; print(json.load(sys.stdin)['document_id'])")

# Wait for processing
for i in {1..30}; do
  STATUS=$(curl -s "http://localhost:9000/v1/documents/$DOC_ID" | \
    python3 -c "import sys,json; print(json.load(sys.stdin)['status'])")
  [ "$STATUS" = "completed" ] && break
  sleep 2
done

# Query document
curl -s -X POST "http://localhost:9000/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -d "{
    \"model\": \"qwen2.5-14b\",
    \"messages\": [{\"role\": \"user\", \"content\": \"分析这份文档的主要内容\"}],
    \"project_id\": \"temp\",
    \"enable_rag\": true
  }"

# Delete document
curl -s -X DELETE "http://localhost:9000/v1/documents/$DOC_ID"

Important Notes

1. No authentication required - Local service, no tokens needed
2. Always use project_id: "temp" for temporary analysis
3. Wait for processing - Documents need 5-30 seconds for OCR/chunking
4. Check chunk_count - If 0, the file may be empty or image-based
5. Always enable_rag: Set to true for document-based answers

Troubleshooting

Issue	Solution
Connection refused	Check: docker ps | grep lihuo
Processing timeout	Large files take longer, wait up to 60s
Empty response	Check if chunk_count > 0
Parse error	File may be corrupted or password-protected