Parallel Agent Dispatch
Split several independent problems across separate AI agents that work at the same time
- Repository
- jnMetaCode/superpowers-zh
- Stars
- 8,444
- Licence
- MIT
- Files
- 1 file
What it does
When you have two or more unrelated tasks, such as several failing test files caused by different bugs, this skill tells the AI to hand each one to its own sub-agent with a focused brief and an isolated context, so they run concurrently instead of one after another. It also guides the AI to write clear, self-contained instructions for each sub-agent, set constraints like not touching other code, and afterwards review each summary, check for conflicts, and run the full test suite. It is a workflow guide written in Chinese; it contains no executable code of its own.
Great for
- Developers facing a batch of unrelated test failures after a large refactor
- Teams whose AI tool supports spawning sub-agents and wants to speed up independent fixes
- Anyone who wants guidance on writing focused, self-contained agent prompts
Skip it if
- Related failures that likely share one root cause, since fixing them separately wastes effort and can conflict (the skill itself says to investigate these together)
- Exploratory debugging where you do not yet know what is broken
- Tasks where agents would edit the same files or shared resources, which can cause interference and conflicting changes
Try asking
- “I have 4 test files failing after the refactor with different causes; dispatch one agent per file to fix them in parallel.”
- “Split these three independent bugs in different subsystems across separate agents and ask each for a summary of root cause and changes.”
- “Write a focused, self-contained prompt for an agent to fix the failing tests in one file without modifying production code.”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
dispatching-parallel-agents.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip dispatching-parallel-agents.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip dispatching-parallel-agents.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 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
- Give each sub-agent one clearly scoped problem, a specific goal, and explicit constraints such as only fixing tests.
- Ask every sub-agent to return a summary of root cause and changes, then check for overlapping edits and run the full test suite before accepting results.
- Use this only when tasks are truly independent; for related failures, investigate together first.
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.