Foundry analyzes your codebase, asks about your workflow, and designs a custom team of specialized agents. One experience if you're building solo, another if you're configuring a team with standards and compliance needs.
Foundry adapts its setup experience and what it generates based on how you work.
A streamlined setup that stays out of your way. The interview asks about your personal workflow and pain points, and Foundry builds a lean team that matches how you actually work.
Upload your policy documents and Foundry extracts enforceable rules from them. Every generated agent, hook, and permission traces back to the source doc that required it.
The building blocks — and the intelligence that ties them together.
The proposal engine reads your code, asks focused questions, and designs agents matched to your stack and conventions — not generic templates.
A local semantic graph of your codebase, updated on every commit. Agents query it instead of reading raw files — ~71× fewer tokens. Free forever, runs entirely offline.
Architectural decisions, conventions, and pain points accumulate across sessions. The Historian files new knowledge; every agent can search it.
Plan-before-acting, conventional commits, protection warnings, standing-order gates. Presets bundle sensible defaults; per-toggle overrides let you fine-tune.
Every agent has a role (Specialist or Guardian) and a kind (Worker or Scholar). Consistent mental model across the roster — visible in every surface.
Everything Foundry generates is standard Claude Code configuration — plain .md files, hooks, settings.json. No lock-in.
.claude/.Free to download and use during the beta period. If you find it useful, donations on Ko-fi help fund development — and beta donors will get credit toward the purchase price at launch.