Self-host a team of AI agents that don't just chat — they work: in a sandbox, on your stack, with your own LLM provider.

Metis is an open, self-hostable agent platform. A real agent with shell and coding tools, reaching your systems through connectors — sandboxed, multi-user, on the provider you choose. Coding is one capability, not the boundary.

Run board — every team's work in flight	Live streaming chat with the agent

• Agents that act, not just answer. Shell, file tools, and connectors (GitHub, Google, Linear, Metabase, …) — the agent reads, runs, edits, and hands back artifacts, sandboxed by default.
• Multi-user from day one. Every resource belongs to a team, and the agent acts as the right person against the right systems — credentials per team and per member. A personal account is just a team of one.
• Your provider, your infrastructure. Anthropic, OpenAI, Google, or whatever comes next, picked per conversation. No Metis-hosted inference, no per-user key vault.
• Build your own tools, and share them. Author skills in the UI (or have the agent write them) and share them across your devices and your team.

Not hypothetical: Metis's sibling, Themis, runs daily in three languages at a company managing 1,000+ properties — one platform, every role:

• Engineering — review a PR for the landmines green tests sail past (a locking migration, a hardcoded credential, a fee missing from the total), then file it as a Linear ticket.
• Operations — run the rental operation across countries: OTA cancellations and calendar sync, pricing schedules, utilities and tax.
• Customer service — draft discount emails, chase a booking anomaly, check whether an inbound email is genuine.
• Leadership — audit dashboard metrics, pull KPI rankings, prep the team-meeting agenda.

Themis is Claude-only (Claude Agent SDK); Metis is the open, provider-agnostic form of the same shape, built on pi, a fast, open agent harness. The lessons come along; the lock-in does not.

pi runs on your laptop. Metis runs pi for everyone you work with — in a sandbox, on your stack, with your provider.

• VISION.md — what Metis is, the rules we hold to, what we won't build.
• PLAN.md — current status, roadmap, and open questions.
• docs/architecture.md — how a turn flows; the Agent service layer.
• docs/configuration.md — runtimes, providers, and environment.

• Rails 8.1, Ruby 4.0.5, PostgreSQL
• pi agent harness, driven via the pi-agent-rb gem
• Any LLM provider pi supports, chosen per conversation
• Hotwire (Turbo + Stimulus, importmap) and Tailwind for the live chat UI
• Devise for auth; Solid Queue / Cache / Cable for jobs, cache, and Action Cable

First, copy the env template and set at least one provider key (e.g. ANTHROPIC_API_KEY) so the agent can talk to a model:

cp .env.example .env   # then edit it: set ANTHROPIC_API_KEY (or another provider)

Then start it with Docker — nothing local to install but Docker. compose.yaml bakes Ruby, pi, the MCP bridge, and the gws CLI into the dev image, runs Postgres alongside, and serves the agent on the local runtime inside the container at http://localhost:3000.

docker compose up --build   # first run, or after a Gemfile change
docker compose up           # subsequent runs

See docs/configuration.md for runtimes, providers, and every variable.

bin/rails test   # full Minitest suite
bin/rubocop      # lint (rubocop-rails-omakase)
bin/ci           # rubocop, security scans, and tests

See docs/architecture.md and CLAUDE.md for architecture and conventions.