opencode-orchestrator-creator

Multi-agent coordination uses HTTP API calls to manage discovery, session creation, and message routing across servers. This Skill automates that setup by creating an orchestrator with curl permissions that discovers available agents via /config and /agent endpoints, then delegates and tracks tasks across your swarm without manual intervention.

Dynamic server discovery automatically detects running OpenCode servers and their capabilities at runtime instead of hardcoding agent addresses. The orchestrator queries endpoints to find available agents and their functions, eliminating configuration maintenance when servers are added or removed.

Use this Skill to initialize a primary agent with curl permissions in your current directory. It enforces a workflow of discovery, session creation, execution, and inter-agent communication, enabling the orchestrator to manage any OpenCode swarm through HTTP requests without agent-specific customization.

Yes. Curl-based HTTP orchestration handles discovery of agents and servers, creates sessions, sends messages, and coordinates task execution across swarms. This approach works universally across any OpenCode deployment without requiring special client libraries or hardcoded server lists.

Enforced introduction formats standardize how agents identify themselves and communicate with the orchestrator and each other. Structured inter-agent protocols reduce misunderstandings, ensure consistent message handling, and enable the orchestrator to delegate tasks reliably across heterogeneous swarms.

No. Universal orchestration eliminates hardcoding by using runtime discovery instead of static configuration. The orchestrator dynamically finds agents, learns their capabilities, manages sessions, and coordinates interactions—so you can deploy agents of different types without reconfiguring the orchestrator.