Tool-Agnostic AI Agent Framework

Set up safe and reliable AI agent workflows in any repository

Start with universal AGENTS.md defaults. Add project context in docs/. Extend behavior with skills and workflows.

Layer 1: Safe Defaults

Set up an AGENTS.md baseline that improves reliability from day one and works across AI coding tools.

Layer 2: Knowledge Base

Add domain terms, architecture decisions, and constraints so agents stop guessing project context.

Layer 3: Skills & Workflows

Capture team-specific expertise and repeatable processes in reusable prompts that scale with your codebase.