Home

BMLibrarian Wiki

Welcome to the BMLibrarian documentation! BMLibrarian is a comprehensive Python library providing AI-powered access to biomedical literature databases with a sophisticated multi-agent architecture.

🚀 Quick Links

For Users

• Getting Started - Installation and setup guide
• CLI Guide - Interactive command-line interface
• Research GUI Guide - Desktop research application
• Configuration Guide - System configuration and settings
• Workflow Guide - Complete research workflow walkthrough
• Multi-Model Query Guide - Using multiple AI models for better search results

For Developers

• Architecture - System architecture overview
• Agent Module - Multi-agent system design
• GUI Architecture - Desktop application architecture
• Multi-Model Architecture - Multi-model query generation internals
• Contributing - How to contribute to the project

📚 Documentation Structure

User Guides

Complete guides for using BMLibrarian's features:

• Getting Started - First-time setup
• CLI Guide - Command-line interface
• Research GUI Guide - Desktop application
• Config GUI Guide - Configuration interface
• Configuration Guide - Settings and options
• Workflow Guide - Research workflow
• Agents Guide - Understanding AI agents
• Query Agent Guide - Natural language queries
• Citation Guide - Citation extraction
• Reporting Guide - Report generation
• Counterfactual Guide - Evidence analysis
• Multi-Model Query Guide - Multi-model search
• Queue Management Guide - Task queue system
• Migration System - Database migrations
• Troubleshooting - Common issues and solutions

Developer Documentation

Technical documentation for contributors:

• Architecture - System design
• Agents Architecture - Multi-agent patterns
• Agent Module - Agent implementation
• GUI Architecture - Desktop GUI design
• Modular CLI Architecture - CLI structure
• Workflow System - Workflow orchestration
• Queue System Architecture - Task queuing
• Citation System - Citation extraction
• Reporting System - Report generation
• Counterfactual System - Evidence analysis
• Multi-Model Architecture - Multi-model queries
• Human Edit Logging - Edit tracking
• Unified Document Card - UI components
• API Reference - API documentation
• Contributing - Contribution guide

🎯 Key Features

Multi-Agent AI System

• QueryAgent - Natural language to PostgreSQL query conversion
• DocumentScoringAgent - Relevance scoring (1-5 scale)
• CitationFinderAgent - Passage extraction
• ReportingAgent - Report synthesis
• CounterfactualAgent - Contradictory evidence search
• EditorAgent - Comprehensive report editing

Advanced Capabilities

• Multi-Model Query Generation - 20-40% better document retrieval
• Task Queue Orchestration - Memory-efficient batch processing
• Human-in-the-Loop - Interactive decision points
• PDF Management - Full-text access and batch downloads
• Counterfactual Analysis - Systematic contradictory evidence search
• Evidence Validation - AI-powered citation verification

Production Features

• PostgreSQL + pgvector - Semantic search database
• Local LLM (Ollama) - Privacy-preserving AI
• GUI Applications - Desktop interfaces
• Comprehensive Testing - >95% test coverage
• Migration System - Automated schema updates

🔗 External Resources

• GitHub Repository
• Issue Tracker
• Pull Requests

📝 Version Information

Current stable version: v0.6

See Release Notes for changes in each version.

---

This wiki is maintained alongside the BMLibrarian codebase. All documentation is also available in the doc/ directory of the repository.