TODO List: AI-Powered Personal Web History Knowledge Graph

This document outlines the implementation tasks required to build the AI-Powered Personal Web History Knowledge Graph application as defined in the PRD and technical specification.

Milestone 1: Project Setup & File Parsing

• Create Tauri project structure with Svelte frontend
• Set up SQLite database with initial schema
• Implement SQLite extractor for Safari history.db files
  • Parse visit data (URL, title, timestamps, visit count)
  • Normalize data from multiple history files
  • Handle error cases and validation
• Design and implement data models for:
  • Visit
  • URL
  • Metadata
• Create file upload interface for history.db files
• Implement source tracking for multiple devices

Milestone 2: Enrichment Pipeline

• Set up OpenAI-compatible API client
  • Configure for both cloud and local endpoints
  • Implement environment-based configuration
• Develop content fetching module for URL data
  • Handle various error cases and timeouts
  • Implement HTML content extraction
• Create AI enrichment pipeline:
  • Summarization functionality
  • Keyword extraction
  • Embedding generation
• Set up local Qdrant vector store
  • Create "url_embeddings" collection
  • Define vector schema with payload structure
• Build batch processing system for URL enrichment
  • Skip already processed URLs
  • Cache failures for retry
  • Implement async batch operations

Milestone 3: Search & UI Implementation

• Design and implement search interface
  • Keyword-based search
  • Semantic search using embeddings
  • Combined search with scoring mechanism
• Create filtering system:
  • Time-based filters
  • Domain filters
  • Topic/tag filters
• Implement search results view with:
  • Link preview capabilities
  • Quick summaries
  • Metadata display

Milestone 4: Dashboard & Timeline Views

• Implement timeline view
  • Chronological display of browsing history
  • Annotation capabilities
  • Color-coding by topic/domain
• Build time series dashboard
  • Daily/weekly visit trends
  • Topic clusters over time
  • Domain distribution charts
• Create thematic clustering visualization
  • NLP-based topic modeling
  • Interactive topic exploration
  • Cluster visualization

Milestone 5: Offline Mode & Local AI Support

• Implement fully offline processing option
  • Local embedding models
  • Local summarization
• Create desktop agent for scheduled syncing (macOS service)
  • Background process for Safari DB monitoring
  • Incremental update system
• Add privacy controls and transparency features
  • Data usage explanations
  • Processing options configuration
  • Data retention settings

Milestone 6: Packaging and User Testing

• Optimize app performance
  • Ensure < 200MB app size
  • < 1GB storage for 50k URLs
• Package application for distribution
  • macOS build process
  • Code signing and notarization
• Create user documentation
  • Setup guide
  • Usage tutorials
  • FAQ section
• Conduct user testing
  • Measure time to first insight
  • Track URL enrichment success rate
  • Gather user satisfaction metrics for search features

Future Enhancements (Post-MVP)

• Implement bookmarking and annotation system
• Add integration with read-later services (Pocket/Instapaper)
• Develop mobile-friendly dashboard
• Create browser extension for real-time data collection

---

Note: This TODO list is based on the Product Requirements Document (PRD-history-graph.md) and technical specification (spec-001-history-graph.md). Tasks should be addressed in milestone order, with each milestone completed before moving to the next.