Sylva 🌲
Auto-generate comprehensive AGENTS.md for your repository using Ax-LLM.
Sylva recursively reads your codebase — folder structures, code conventions, state management patterns, architecture styles — and distills it into a standardized AGENTS.md manifest that AI coding assistants (and human engineers) can follow perfectly.
Why Sylva?
Modern AI coding tools (Cursor, Claude Code, Copilot, Antigravity) perform dramatically better when given explicit project context. Sylva automates the creation of that context by:
- Serializing your entire source tree into a structured format
- Analyzing it with a Reasoning Language Model (RLM) agent
- Compiling the analysis into a standardized 17-section AGENTS.md
Quick Start
# Run instantly on any local repository
npx @thelogicatelier/sylva --local-repository /path/to/your/project -m openai/gpt-4o -i 5
# Or analyze a GitHub repository directly
npx @thelogicatelier/sylva --github-repository https://github.com/expressjs/express
What You'll Find Here
| Section | Description |
|---|---|
| Getting Started | Installation, configuration, and your first run |
| Core Concepts | How the pipeline works under the hood |
| Models & Providers | Supported LLMs and how to choose the right one |
| CLI Reference | Every flag, environment variable, and exit code |
| Output Format | What AGENTS.md contains and how to customize it |
| Advanced Usage | Monorepos, CI/CD, and troubleshooting |
| Architecture | Internal code architecture for contributors |
| Contributing | Development setup, code quality, and PR guidelines |
Built by The Logic Atelier