ACE helps teams scale judgment, not just tooling. It turns conventions, project context, and
hard-won technical experience into a working coding environment so every session can start from
what the team already knows, not from whatever one developer configured last.
Built for CTOs, product owners, and engineering managers who want AI tooling to reflect how
their teams actually build software, not just how one developer happened to set up one machine.
// why teams use it
Make senior engineering judgment available across the team instead of trapped in a few people or repos
Bring product direction, technical standards, and project context into day-to-day coding work
Reduce setup drift by loading shared skills, MCP tools, and backend context from one source
Support different roles from the same foundation without forcing everyone into the same way of working
// quick start
# macOS / Linux $ curl -fsSL https://raw.githubusercontent.com/prod9/ace/main/install.sh | bash # Windows (PowerShell) > powershell -c "irm https://raw.githubusercontent.com/prod9/ace/main/install.ps1 | iex"
$ ace setup prod9/school # clone a school, register MCP, write config $ ace # launch your AI backend
// supported backends
Claude CodeCodex
// what a school gives you
A shared home for prompts, coding skills, conventions, MCP endpoints, and project context
A practical way to spread your strongest technical practices beyond the people who already know them
A feedback loop where teams improve shared skills once and reuse those improvements across projects
// how it works
1. ACE resolves the active school, backend, and project settings for the current repository
2. It clones or updates the school cache and links shared folders into the backend workspace
3. It registers the MCP tools the team expects to have available
4. It composes session context from school guidance and project config
5. Then it hands off to Claude Code or Codex with the environment ready to work