Skip to content

Latest commit

 

History

History
107 lines (81 loc) · 2.44 KB

commands.md

File metadata and controls

107 lines (81 loc) · 2.44 KB

Quick Reference - Vector Bot

Essential Commands

# System check
vector-bot doctor

# Index documents
vector-bot ingest

# Query documents
vector-bot query "your question"

# Show configuration
vector-bot --config-info

File Structure

your-project/
├── vector-bot.exe           # Executable
├── docs/             # Your documents go here
├── index_storage/    # Generated index (don't edit)
└── .env              # Optional config file

Document Types Supported

  • PDF files (.pdf)
  • Text files (.txt)
  • Markdown (.md)
  • JSON files (.json)
  • CSV files (.csv)

Common Usage Patterns

First Time Setup

ollama pull llama3.1                    # Install AI model
ollama pull nomic-embed-text            # Install embedder
vector-bot doctor                              # Check setup

Daily Workflow

cp new-document.pdf docs/               # Add document
vector-bot ingest                              # Update index
vector-bot query "What's new in this doc?"     # Ask question

Advanced Queries

vector-bot query "summary?" --k 8              # More context
vector-bot query "details?" --show-sources     # Show sources

Environment Variables

Variable Default Purpose
DOCS_DIR ./docs Where documents are stored
OLLAMA_CHAT_MODEL auto Which AI model to use
SIMILARITY_TOP_K 4 Context chunks to retrieve

Troubleshooting

Problem Solution
"Server not running" ollama serve
"No models" ollama pull llama3.1
"No documents found" Add files to docs/
"Index not found" Run vector-bot ingest

Development Commands

# Run tests
pytest tests/ -v
python run_tests.py

# Test coverage
pytest tests/ --cov=rag

# Code quality
ruff check src/
mypy src/

# Security scan
safety check
bandit -r src/

Quick Tips

  • Add documents: Drop files in docs/ folder
  • Re-index: Run vector-bot ingest after adding files
  • Better answers: Try --k 6 for more context
  • Find sources: Use --show-sources flag
  • Check config: Use vector-bot --config-info
  • Run tests: Use pytest tests/ or python run_tests.py
  • Contribute: See CONTRIBUTING.md

Full documentation: USER_GUIDE.md | Testing: TESTING.md