Gencodedoc

🚀 GenCodeDoc

Smart documentation generator and intelligent versioning system with full MCP support

*Intelligent versioning and documentation for modern development workflows*

Quick Start • Features • MCP Tools • Documentation

---

⚡ Quick Start

# 1. Install

cd /path/to/gencodedoc poetry install

2. Initialize your project

poetry run gencodedoc init --preset python

3. Create your first snapshot

poetry run gencodedoc snapshot create -m "Initial version" -t v1.0

4. Generate documentation (Smart Split)

poetry run gencodedoc doc generate --limit 5000

5. Visualize Tree

poetry run gencodedoc tree

🎯 For AI Assistants (Claude/Gemini): See MCP Integration

✨ Features 🎯 Core Features 📸 Smart Snapshots - Create intelligent snapshots with ~70% space savings via SHA256 deduplication 🔄 Intelligent Autosave - 3 modes (timer/diff/hybrid) with configurable thresholds 📝 Beautiful Documentation - Generate Markdown docs with syntax highlighting, directory trees, and smart splitting 🔍 Advanced Diff - Compare versions with unified, JSON, or AST-based diffs 🗜️ Efficient Storage - zstd compression (~3x reduction) + SQLite with optimized indexes 🎨 Project Presets - Pre-configured for Python, Node.js, Go, and Web projects 🔌 MCP Integration (Model Context Protocol) 26 MCP Tools - Full CLI functionality exposed via MCP for AI Assistants 3 Transports - stdio (Gemini CLI) + SSE (Claude Desktop) + REST API Multi-Project - Manage multiple projects simultaneously Code Intelligence - File history, search, and changelog generation Live Status - Real-time project statistics and snapshot management 📦 Installation Prerequisites Python 3.10+ Poetry (dependency manager) Install Poetry Bash

Linux/macOS

curl -sSL https://install.python-poetry.org | python3 -

Windows (PowerShell)

(Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | py - Install GenCodeDoc Bash

Clone or navigate to the project

cd /home/fkomp/Bureau/oracle/utilitaires/gencodedoc/gencodedoc

Install dependencies

poetry install

Verify installation

poetry run gencodedoc --help 🎯 Usage 📋 CLI Usage Initialize a Project Bash

Basic initialization

gencodedoc init

With Python preset (recommended for Oracle projects)

gencodedoc init --preset python

Available presets: python, nodejs, go, web

Snapshot Management Bash

Create snapshot

gencodedoc snapshot create --message "Feature X completed" --tag v1.0

List snapshots

gencodedoc snapshot list --limit 10

Show snapshot details

gencodedoc snapshot show v1.0

View file content at specific version (NEW)

gencodedoc snapshot cat v1.0 src/main.py

List files in a snapshot (NEW)

gencodedoc snapshot files v1.0 --pattern "*.py"

Compare versions

gencodedoc snapshot diff v1.0 v2.0

Restore snapshot (full or partial)

gencodedoc snapshot restore v1.0 --force gencodedoc snapshot restore v1.0 --filter "src/*.py" # Partial restore

Export snapshot (NEW)

gencodedoc snapshot export v1.0 ./dist-v1 --archive # Creates .tar.gz

Delete snapshot

gencodedoc snapshot delete old-version --force

Cleanup orphaned data (NEW)

gencodedoc snapshot cleanup Documentation Generation Bash

Full documentation

gencodedoc doc generate

Custom output

gencodedoc doc generate --output docs/API.md

Specific paths only

gencodedoc doc generate --include src/api/ --include README.md

Structure preview

gencodedoc doc preview --max-depth 3

Project statistics

gencodedoc doc stats Configuration Bash

View configuration

gencodedoc config show

Edit configuration

gencodedoc config edit

Set specific values

gencodedoc config set autosave.enabled true gencodedoc config set autosave.mode hybrid

Apply preset

gencodedoc config preset python

Manage ignore rules

gencodedoc config ignore --add-dir dist gencodedoc config ignore --add-ext .tmp gencodedoc config ignore --list-all

Customizable Presets

Presets are now defined in YAML files (gencodedoc/config/presets/*.yaml)

and can be modified directly in the source code.

Project Status Bash

Show project status

gencodedoc status 🔌 MCP Integration GenCodeDoc exposes 22 powerful tools via the Model Context Protocol, compatible with Claude Desktop, Gemini CLI, and any MCP-compatible client.

🎯 Transport Modes Transport Use Case AI Assistants Port stdio CLI integration Gemini CLI, custom scripts stdin/stdout SSE Web/Desktop apps Claude Desktop, web UIs 8000 (HTTP) REST API integration Any HTTP client 8000 (HTTP) 🚀 Setup for Gemini CLI (stdio)

1. Find your Poetry venv path:

Bash

cd /home/fkomp/Bureau/oracle/utilitaires/gencodedoc/gencodedoc poetry env info --path

Copy the path and append /bin/python

1. Add to ~/.config/gemini-desktop-app/settings.json:

JSON

{ "mcpServers": { "gencodedoc": { "command": "/path/to/your/venv/bin/python", "args": ["-m", "gencodedoc.mcp.server_stdio"], "env": { "PROJECT_PATH": "/home/fkomp/Bureau/oracle/utilitaires/gencodedoc/gencodedoc" } } } }

1. Restart Gemini CLI and you`re ready! 🎉

🚀 Setup for Claude Desktop (SSE)

1. Start the SSE server:

Bash

Terminal 1: Start server

poetry run python -m gencodedoc.mcp.server_sse

Server runs on http://127.0.0.1:8000

1. Add to Claude Desktop config:

Location:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json Linux: ~/.config/Claude/claude_desktop_config.json Config:

JSON

{ "mcpServers": { "gencodedoc": { "url": "http://127.0.0.1:8000/mcp/sse", "transport": "sse", "description": "GenCodeDoc - Smart documentation and versioning" } } }

1. Restart Claude Desktop and the server must remain running!

🚀 REST API Bash

Start REST server

poetry run python -m gencodedoc.mcp.server

Available endpoints

GET http://127.0.0.1:8000/ # Server info GET http://127.0.0.1:8000/mcp/tools # List tools POST http://127.0.0.1:8000/mcp/execute # Execute tool 🛠️ MCP Tools (26 Tools)

🧠 Code Intelligence (3 tools)

📸 Snapshot Management (11 tools)

📝 Documentation (3 tools)

🎯 Project Management (2 tools)

⚙️ Configuration (4 tools)

🔄 Autosave (3 tools)

---

🤝 Contributing Contributions welcome! 🎉

🐛 Report bugs via GitHub Issues 💡 Suggest features via Discussions 🔧 Submit Pull Requests

Development workflow:

1. Fork the repo
2. Create a feature branch (git checkout -b feature/amazing)
3. Make your changes
4. Run tests (make test)
5. Submit PR

📄 License MIT License - See LICENSE file

🐛 Issues: GitHub Issues 💬 Discussions: GitHub Discussions 🌐 Website: esprit-artificiel.com 📧 Email: support@esprit-artificiel.com

Made with ❤️ for developers who love smart versioning and beautiful docs

💡 Pro Tip: Use GenCodeDoc with Gemini CLI or Claude Desktop for AI-powered version control!

⭐ Star this repo if you find it useful! 🤖 "Start autosave in hybrid mode for this project" → Calls: start_autosave(project_path="...", mode="hybrid")

🤖 "Stop all autosaves and show me their status" → Calls: stop_autosave(...) + get_autosave_status() ⚙️ Configuration Configuration File (.gencodedoc.yaml) YAML

project_name: "my-project"

Files/directories to ignore

ignore: dirs:

• node_modules
• venv
• .venv
• pycache
• .git
• dist
• build files:
• "*.log"
• package-lock.json
• .DS_Store extensions:
• .pyc
• .pyo
• .exe
• .jpg
• .png
• .pdf patterns: [] # gitignore-style patterns

Intelligent autosave

autosave: enabled: false # Set to true to enable mode: hybrid # timer | diff | hybrid (recommended)

Timer mode: fixed interval

timer: interval: 300 # seconds (5 minutes)

Diff mode: threshold-based

diff_threshold: threshold: 0.05 # Save if 5% of files changed check_interval: 60 # Check every 60 seconds ignore_whitespace: true ignore_comments: false

Hybrid mode: combines timer + diff (RECOMMENDED)

hybrid: min_interval: 180 # Minimum 3 minutes between saves max_interval: 600 # Maximum 10 minutes between saves threshold: 0.03 # Save if 3% of files changed

Retention policy

retention: max_autosaves: 50 # Keep max 50 autosaves compress_after_days: 7 # Compress after 7 days delete_after_days: 30 # Delete after 30 days keep_manual: true # Always keep manual snapshots

Documentation output

output: default_name: "{project}_doc_{date}.md" include_tree: true include_code: true tree_full_code_select: false # Full tree, selected code only language_detection: true max_file_size: 1000000 # 1 MB max per file

Diff output format

diff_format: default: unified # unified | json |