v1.0.0 - Production Ready

tg-parser v1.0.0 - Production Ready 🎉

First stable release! Parse Telegram Desktop JSON exports for LLM processing.

✨ Features

• Parse all chat types: Personal, groups, supergroups, forum topics, channels
• Advanced filtering: 9 filter types (date, sender, topic, content, attachments, reactions, forwards, service messages, composite)
• Smart chunking: 3 strategies for LLM context limits (fixed, topic, hybrid)
• Streaming support: Handle large exports (>50MB) with automatic ijson-based streaming
• Multiple formats: Markdown (LLM-optimized), JSON, KB-template with YAML frontmatter
• MCP integration: 6 tools for Claude Desktop/Code
• Production quality: 261 tests, pyright strict mode, Clean Architecture

📦 Installation

```bash

From PyPI

pip install tg-parser

With uv (recommended)

uv tool install tg-parser

From source

git clone https://github.com/mdemyanov/tg-parser.git
cd tg-parser
uv sync --all-extras
```

🚀 Quick Start

```bash

Parse chat export

tg-parser parse ./export.json

Filter by date and sender

tg-parser parse ./export.json --last-days 7 --senders "Alice,Bob"

Chunk for LLM context

tg-parser chunk ./export.json -s hybrid --max-tokens 8000

Get statistics

tg-parser stats ./export.json
```

📚 Documentation

• README - Installation and usage
• ARCHITECTURE - Clean Architecture design
• DEVELOPMENT - Development guide
• TELEGRAM_FORMAT - Telegram JSON spec
• PRD - Product requirements and roadmap
• CHANGELOG - Version history

🙏 Credits

Built with Claude Code following Clean Architecture principles.

See CHANGELOG.md for full version history.