# AI Agent Instructions for Tenets
This document helps AI agents understand the Tenets codebase.
## Quick Reference
### What is Tenets?
MCP server, CLI, and Python library for intelligent code context. Helps AI assistants understand codebases.
### Key Entry Points
```bash
# CLI
tenets distill "description" # Build code context
tenets examine # Analyze codebase
# MCP Server
tenets-mcp # Start MCP server
```
### Key Files
| File | Purpose |
|------|---------|
| `tenets/__init__.py` | Main `Tenets` class and public API |
| `tenets/mcp/server.py` | MCP server with tools/resources/prompts |
| `tenets/cli/app.py` | CLI commands |
| `tenets/core/ranking/` | File ranking algorithms |
| `tenets/models/context.py` | `ContextResult` data model |
### Python API
```python
from tenets import Tenets
# Initialize
t = Tenets(path="/project")
# Build context
result = t.distill("find authentication code")
print(result.markdown)
# Create session
t.session_create("my-session")
t.session_pin_file("src/auth.py", "my-session")
# Add guiding principles
t.tenet_add("Always validate user input", priority="high")
```
### MCP Tools
| Tool | Use For |
|------|---------|
| `distill` | Build context for coding tasks |
| `rank_files` | Preview relevance without content |
| `examine` | Codebase structure analysis |
| `session_create` | Start stateful session |
| `session_pin_file` | Pin important files |
| `tenet_add` | Add coding guidelines |
## Development
### Setup
```bash
pip install -e ".[dev,mcp]"
```
### Testing
```bash
pytest tests/
```
### Linting
```bash
ruff check tenets/
mypy tenets/
```
## Architecture
```
tenets/
├── __init__.py # Tenets class
├── mcp/
│ └── server.py # TenetsMCP class
├── core/
│ ├── nlp/ # Text processing
│ └── ranking/ # Relevance scoring
├── models/
│ └── context.py # ContextResult
└── cli/
└── app.py # Typer CLI
```
## Contact
- GitHub: https://github.com/jddunn/tenets
- Email: team@manic.agency
