Zero-install, files-only development framework for AI coding agents. SCOPE → BUILD → SHIP.
# KEEL **The structural backbone. First thing laid down. Everything builds on it.** KEEL is a zero-install, files-only development framework for AI coding agents. No CLI. No package manager. No update treadmill. Copy the files, start building. ## Why KEEL Exists AI coding agents degrade over long sessions. As the context window fills up, output quality drifts — instructions get missed, code gets sloppy, details get lost. This is context rot, and it's the single biggest reliability problem in AI-assisted development. KEEL solves it with files. The framework IS the files. A `CLAUDE.md` that contains every instruction your AI agent needs, and a `docs/` directory that holds project state as human-readable markdown. When you clear context and start a new session, the agent reads the files and picks up exactly where it left off. No state is lost because no state lived in memory — it was always on disk. No runtime. No dependencies. No config. If your agent can read markdown, KEEL works. ## Quick Start ```bash git clone https://github.com/diydigitaldreams/keel.git cp -r keel/template/.cursor your-project/.cursor cp -r keel/template/docs your-project/docs ``` Or use the setup script: ```bash git clone https://github.com/diydigitaldreams/keel.git ./keel/setup.sh /path/to/your-project ``` Then open your AI coding agent and say: > "Let's scope this project." KEEL takes it from there. ## How It Works Three modes. No ceremony. ``` SCOPE → Interview, research, spec out what we're building BUILD → Plan atomic tasks, execute, commit each one SHIP → Verify success criteria, log, advance to next phase Repeat BUILD → SHIP for each phase until the project is done. ``` Context rot is eliminated by clearing context between phases. Each phase gets its own directory with only the files that phase needs. Heavy thinking — research, planning, architectural decisions — happens once and gets written to disk. Every future context window inherits that work by reading the
Agent that generates comprehensive documentation, API references, architecture diagrams, and developer onboarding guides from existing code.
Agent configuration for systematic bug investigation that traces issues from error logs through the codebase to root cause with suggested fixes.
Agent for integrating third-party APIs including SDK setup, type generation, error handling, retry logic, and rate limit management.
Cursor's built-in autonomous coding agent that can make multi-file edits, run terminal commands, search the codebase, and iteratively build features with minimal human intervention.
Cloud-based autonomous coding agent that runs in the background on remote sandboxed environments, handling complex multi-step tasks while you continue working.
Cursor's multi-file editing agent within Composer mode that can create, edit, and delete files across your entire project in a single conversation.