AGENT-33

Autonomous AI agent orchestration engine with a local-first runtime, explicit governance, and extensible workflow automation.

Repository Layout

• engine/: FastAPI runtime, API routes, orchestration services, tests
• frontend/: first-party AGENT-33 control plane UI
• core/: orchestration specifications, protocol references, policy packs
• docs/: canonical operational documentation for current runtime behavior

Quick Start

cd engine
cp .env.example .env
docker compose up -d
curl http://localhost:8000/health

If reusing an Ollama container from another compose project (no host port mapping), start with the shared-network override:

docker compose -f docker-compose.yml -f docker-compose.shared-ollama.yml up -d

Open the control plane UI:

• http://localhost:3000

Default local credentials (from .env.example):

• username: admin
• password: admin

⚠️ Security Warning: The bootstrap authentication (admin/admin) is for local development only and must not be used in production environments. For production or VPS deployments, disable bootstrap auth (AUTH_BOOTSTRAP_ENABLED=false) and configure a proper identity provider or token issuing mechanism.

Generate a local development JWT (1 hour):

docker compose exec -T api python -c "import os,time,jwt; now=int(time.time()); payload={'sub':'local-admin','scopes':['admin','agents:read','agents:write','agents:invoke','workflows:read','workflows:write','workflows:execute','tools:execute'],'iat':now,'exp':now+3600}; print(jwt.encode(payload, os.getenv('JWT_SECRET','change-me-in-production'), algorithm=os.getenv('JWT_ALGORITHM','HS256')))"

Call a protected endpoint:

export TOKEN="<paste-token-here>"
curl http://localhost:8000/v1/agents/ -H "Authorization: Bearer $TOKEN"

Documentation

Start here:

• Documentation Index
• Setup Guide
• Walkthroughs
• Use Cases
• Functionality and Workflows
• API Surface
• PR Review (2026-02-15)

Reference material:

• engine/docs/
• core/
• docs/phases/

Current Status

The project is under active development on the unified UI platform branch. Core capabilities include agent orchestration, workflow execution, memory/RAG, and operational automation (reviews, releases, evaluations, autonomy budgets, improvements).

For a historical snapshot of PR review activity as of February 15, 2026, see docs/pr-review-2026-02-15.md.

License

MIT. See LICENSE.