Skip to content
Get packs

Workflow Runner

Runs multi-role YAML workflows inside your AI tool, playing each role in turn with no API key

Source Download skill
Repository
jnMetaCode/superpowers-zh
Stars
8,444
Licence
MIT
Files
1 file

What it does

When you give the agent a YAML workflow file or ask several expert roles to work together on a task, this skill tells it to read the workflow, check the role definition files, and then act out each role in order, passing each step's output to later steps. The current AI session does all the work, so no separate API key is needed. Results are saved as files in a dated output folder.

Great for

  • People who already use agency-agents-zh role files and want to run multi-step workflows inside Claude Code or a similar tool
  • Users who want a team-style review, such as a product manager and architect critiquing a spec, without setting up a separate orchestration service
  • Chinese-speaking users of agent tools, since the skill and its instructions are written in Chinese

Skip it if

  • Tasks needing guaranteed independent expert opinions: all roles run in one model session, so they can blend together rather than truly disagree
  • Workflows that must use a specific model, timeout or retry settings, because the skill explicitly ignores those YAML fields
  • Anyone who cannot install the agency-agents-zh role files, since the skill stops if the role directory is missing

Try asking

  • “Run workflows/story-creation.yaml with the story idea I provide”
  • “Have a product manager and a software architect review this PRD together”
  • “Use a narrator and a psychologist to write a short story about a lighthouse keeper”
  • “Show me the execution plan first, then run the product-review workflow”
Step-by-step guide: how to use Workflow RunnerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    workflow-runner.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip workflow-runner.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip workflow-runner.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Install agency-agents-zh first, either with git clone --depth 1 or npm install, so the skill can find the role files
  • Review the execution plan and any auto-generated workflow before confirming, since the agent writes both from your description
  • Check the saved steps folder in .ao-output to see each role's individual output, not just the final summary

What's inside

  • SKILL.md

Redistributed unmodified from jnMetaCode/superpowers-zh under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.