distributed-task-execution — Claude Code skill: use and install
The distributed-task-execution skill splits unrelated problems across agents, runs investigations in parallel, and gathers their results. It fits debugging work with separate failures.
LicenseRef-Proprietary 134 installs ★ 50 updated 2026-01-05
npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a claude-codeWhat distributed-task-execution does
The skill turns separate failures into focused agent tasks. Each task gets a defined scope, a goal, limits on what may change, and an expected report. Once the agents finish, you review their summaries, check for overlapping edits, and run the full test suite.
It is meant for cases where failing tests or broken subsystems have separate causes and can be examined without shared state. Related failures should stay together, since one fix may affect the others. Parallel work is also a poor fit when agents edit the same files or the problem still needs broad system exploration.
The skill is available for Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, and OpenCode. In Claude Code, install it with npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a claude-code; the installed file is ~/.claude/skills/distributed-task-execution/SKILL.md.
What it helps with
- split test failures by independent problem domain
- dispatch several investigations at once
- give each agent a bounded debugging task
- review returned summaries before integration
Example prompts for the agent
- Split the failures in
agent-tool-abort.test.ts,batch-completion-behavior.test.ts, andtool-approval-race-conditions.test.tsinto separate agent tasks. - Run parallel investigations for three unrelated failures and have each agent report the root cause and edits.
- Review every agent summary, check for overlapping changes, then run the full test suite.
How to install distributed-task-execution
1Skills CLI (Vercel)
npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a claude-code2Manually: put the files in place
~/.claude/skills/distributed-task-execution/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/distributed-task-execution/
1Skills CLI (Vercel)
npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a codex2Manually: put the files in place
.agents/skills/distributed-task-execution/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a cursor2Manually: put the files in place
.cursor/skills/distributed-task-execution/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install qodex-ai/ai-agent-skills skills/distributed-task-execution2Manually: put the files in place
.github/skills/distributed-task-execution/SKILL.md1Skills CLI (Vercel)
npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a gemini-cli2Platform CLI
gemini skills install https://github.com/qodex-ai/ai-agent-skills --path skills/distributed-task-executionManually: put the files in place
.devin/skills/distributed-task-execution/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about distributed-task-execution
How do I install distributed-task-execution in Claude Code?
Run npx skills add qodex-ai/ai-agent-skills --skill distributed-task-execution -a claude-code, then check ~/.claude/skills/distributed-task-execution/SKILL.md.
What does distributed-task-execution do?
It helps split unrelated debugging work between agents, run those investigations concurrently, and collect their reports. It also describes how to review changes and verify the combined result.
When should I use distributed-task-execution for debugging?
Use it when failures have separate root causes, each area can be understood on its own, and the agents do not share state. Keep related failures or overlapping edits in a sequential workflow.