Back to workflow skills
Development and engineeringLong-running multi-agent frameworkPW
AuthorCommunity author
Development and engineering Skill

Long-running multi-agent framework

Use the Claude SDK to build the Planner→Generator→Evaluator multi-proxy framework, suitable for multi-proxy pipelines, agent loops, and complex orchestration.

0Uses7Likes9Views

What this skill does

Use the Claude SDK to build the Planner→Generator→Evaluator multi-proxy framework, suitable for multi-proxy pipelines, agent loops, and complex orchestration.

  • Clarifies goals, inputs, and success criteria
  • Preserves evidence and marks uncertain boundaries
  • Produces structured, editable results
Installationnpx clawhub@latest install long-run-harnessView source