agent-harness-runtime-charter - Orchestrate delegated agent harness runs
Interprets loaded harness skills and executes them as delegated agent-call graphs with a parent runtime and task children.
Tags
Updated: 2026-09-30Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Interpret harness skills
- Construct execution graphs
- Launch child agents
- Route task stages
- Manage agent-call boundaries
- Close completed runs
Inputs
- Benchmark task
- Loaded harness skills
- Task packets
- Relevant file paths
Outputs
- Agent-call graph
- Child-agent executions
- Task artifacts
- Run conclusions
- Runtime state updates
Requirements
- Access to spawn_agent
- Agent-capable runtime
- Loaded skill context