v0.1.0 — first release

v0.1.0 — first release

The inaugural release of CorpOS, a greenfield reference architecture for an
autonomous company: a TypeScript multi-agent runtime with a policy, approval,
and audit control plane. Every consequential action is policy-gated; nothing
destructive runs autonomously.

Highlights

• Multi-agent runtime — support, sales, ops, finance, and engineering agents
  reason via an LLM, call permissioned tools, and coordinate through a shared
  orchestrator.
• Policy engine as the single chokepoint — allow / deny / approve decisions,
  glob-matched rules, per-task spend caps, and human approval gates.
• Typed tool registry — JSON-schema-validated tools with a permission model
  (read / write / spend / communicate / system / delegate).
• Conversation-preserving approvals — agents pause at a gate and resume from
  the exact state where they paused (no lost context, no re-prompting).
• Control plane — Express REST + WebSocket API and a real-time dashboard
  (live task board, agent statuses, pending approvals, spend meter, event feed).
• One-click demo mode — a curated 5-agent scenario that auto-plays the full
  policy-gated loop.
• Simulation-first — fully demonstrable and tested with no LLM key and no
  network; drop in an OpenRouter key (Owl Alpha) for live reasoning.
• Production-shaped DevOps — CI (Node 20/22), ESLint + Prettier, CodeQL,
  Dependabot, branch protection, and a multi-arch Docker image on GHCR.

Try it

docker run --rm -p 3000:3000 ghcr.io/safelymp/corpos:0.1.0
# → http://localhost:3000  (simulation mode), then click ▶ Run demo

Or one-click deploy via the "Deploy to Render" button in the README.

Security

Reference architecture / research demo — not production-hardened. The control
plane has no authentication, authorization, or rate limiting; do not expose it to
untrusted networks. See SECURITY.md.

Full changelog: https://github.com/SafetyMP/CorpOS/commits/v0.1.0