agent-orchestrate
Mission
Section titled “Mission”Decompose → assign → coordinate → synthesize results. Every orchestration produces a coherent unified output.
When to Use
Section titled “When to Use”ONLY use when explicitly coordinating multiple specialized agents on a shared task, designing multi-agent pipelines, routing tasks between agents, synthesizing results from parallel agents, or managing agent handoffs and context flow.
Do NOT use for single-task requests — use system-design, evidence-research, code-review, or feature-implement instead.
Triggers: “coordinate agents”, “multi-agent workflow”, “agent pipeline”, “assign tasks to agents”, “parallel agents”, “orchestrate this workflow”
Skills Invoked
Section titled “Skills Invoked”orch-agent-orchestrator— decompose request into skill execution planorch-delegation— assign skills to agents/modelsorch-multi-agent— manage concurrent execution, state, partial failuresorch-result-synthesis— merge parallel outputs, surface contradictions
Chain-To
Section titled “Chain-To”quality-evaluate— evaluate the orchestration output qualityfault-resilience— add fault tolerance to the orchestration
Companion Tools
Section titled “Companion Tools”orchestration-config(read/write) — inspect or patch orchestration configurationmodel-discover— list available models and capabilities
Example
Section titled “Example”{ "request": "Run architecture review, security audit, and performance analysis on the new payment service in parallel"}Output: Coordinated results from three parallel agent tasks, contradiction surface, and unified synthesis.