All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

• Analysis Pipeline: Token counting, chunking, salience scoring, and redundancy detection
• Compression Strategies: Deduplication, summarization, sliding window, and hybrid approaches
• Budget Optimization: Greedy and ILP-based chunk selection under token limits
• Tokenization: Support for GPT-4, Claude, and Llama models via tiktoken and sentencepiece
• Embeddings: OpenAI embedding provider with fake provider for testing

• Python SDK: Async API with analyze(), compress(), and optimize() functions
• CLI: Commands for analyze, compress, and viz operations
• REST API: FastAPI-based API with endpoints for analysis, compression, and visualization
• Web UI: SvelteKit dashboard with interactive charts and visualizations

• Storage: SQLite + JSONL persistence for analysis runs
• Testing: Comprehensive test suite with 80%+ coverage
• Documentation: MkDocs site with tutorials and API reference
• CI/CD: GitHub Actions for testing, linting, and releases
• Docker: Containerized API and web applications

• Embedding scatter plots (UMAP projection)
• Token timeline charts
• Interactive run dashboard
• CLI text-based visualizations

• Quick start guide
• Three detailed tutorials
• Complete API reference
• Example notebooks and scripts