What it does
Purpose-built operator interface for coordinating heterogeneous AI agents across separate runtimes. Fills a category gap: no existing framework (AutoGen, LangGraph, CrewAI, Temporal) is designed for human-as-coordinator architecture.
Architecture / Key capabilities
- Human-as-architecture model — The operator is the coordination mechanism, not a passive monitor. Every routing, delegation, and escalation decision flows through structured human judgment
- Heterogeneous runtime support — Coordinates agents running on different platforms (Claude Code, Cowork, Codex, Gemini) without requiring a shared execution environment
- Category gap positioning — Existing frameworks assume agents share a runtime and communicate programmatically. This targets the case where they do not, and a human bridges the gap
- Structured delegation interface — Surfaces agent capabilities, current state, and task fitness so the operator can make informed routing decisions quickly
- Cross-session continuity — Maintains coordination context across operator sessions so multi-day workflows do not lose momentum
Key numbers
MISSING
Current phase
MISSING — would need orchestrator-workstation _docs/current-state.md or design specs to determine
Status
Active
Links
MISSING — Repository URL