workflow-runner

Executes agency-orchestrator YAML workflows using the current session LLM as the engine.

Updated Aug 14, 2026
One-click install
npx skills add https://github.com/bill80362/laravel-thread-admin --skill workflow-runner-bill80362
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: workflow-runner
Source: https://github.com/bill80362/laravel-thread-admin/tree/main/.github/superpowers/workflow-runner
Command: npx skills add https://github.com/bill80362/laravel-thread-admin --skill workflow-runner-bill80362

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Running multi-role agency-orchestrator YAML workflows normally requires a CLI setup with API keys. This Skill lets you execute those workflows directly inside Claude Code, OpenClaw, or Cursor, using the current session's LLM as the execution engine—no API key configuration needed. ## Core Features & Use Cases - YAML Workflow Parsing: Reads workflow files with inputs, steps, roles, and dependency declarations, then builds a topological execution plan. - Multi-Role Orchestration: Loads agent role definitions from an agents directory and plays each role in sequence, running independent steps in parallel via sub-agents. - Structured Output Persistence: Saves each step's output, a final summary, and metadata.json into a dated .ao-output directory. - Use Case: A user provides a story-creation.yaml workflow where a narrative designer drafts a plot and a psychologist refines character motivations; the Skill runs both roles in dependency order and saves the final story. ## Quick Start Run the workflow defined in workflows/story-creation.yaml using the roles from the agency-agents-zh directory.

Frequently Asked Questions about workflow-runner

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run an agency-orchestrator YAML workflow without an API key?

Provide the YAML workflow file path in Claude Code, OpenClaw, or Cursor and ask to run it. The Skill parses the file, loads role definitions, and executes each step using the current session's LLM, so no separate API key is required.

How do I orchestrate multiple AI agent roles in one session?

Define a YAML workflow with steps specifying role paths, tasks, and dependencies. The Skill topologically sorts the steps, plays each role using its markdown definition, and runs independent steps in parallel via sub-agents.

What happens if the agency-agents-zh role directory is missing?

Execution stops and the Skill prompts you to install the role definitions, either by cloning the agency-agents-zh repository with git or installing it via npm. Steps are never executed without first reading the role markdown files.

Can I run a multi-role workflow without writing a YAML file?

Yes. Describe the collaboration you need, such as having a product manager and architect review a PRD, and the Skill generates a YAML workflow definition for your confirmation before executing it.

What happens when a workflow step fails during execution?

The failed step is marked as failed, and all downstream steps that depend on it are skipped. Independent steps continue executing, and results are still saved to the .ao-output directory with per-step status in metadata.json.