run-parallel-agents-feature-build — Claude Code skill: install
The run-parallel-agents-feature-build skill splits independent build work across agents, then checks their results. It is intended for tasks with at least three parallel lanes.
no license 145 installs ★ 5 updated 2026-08-31
npx skills add ulpi-io/skills --skill run-parallel-agents-feature-build -a claude-codeWhat run-parallel-agents-feature-build does
Before launching agents, the skill checks that each build lane can proceed on its own, with separate write scopes and no dependency on another lane’s output. It assigns a suitable agent to each lane, gives every fresh agent a self-contained brief, and starts the work concurrently in isolated worktrees.
Once the agents finish, the skill checks for conflicts and reports each lane’s outcome separately. It handles build work, not research or debugging. Fewer than three independent lanes, shared write scope, or dependencies between lanes rule out a parallel run; those tasks should be handled inline.
You need a concrete task list or an approved plan, along with agents that can take the work. Each agent receives the relevant context in its own brief, since new agents do not inherit the main conversation.
What it helps with
- split a build into at least three independent lanes
- check file overlap and dependencies before starting agents
- give each agent a brief with scope and acceptance criteria
- review parallel results and track blockers by lane
Example prompts for the agent
- Build these three independent parts in parallel: update the API, add a settings screen, and write tests for a separate module.
- Run the independent build tasks from this approved plan in parallel.
- Check whether these tasks can run concurrently without overlapping files.
How to install run-parallel-agents-feature-build
1Skills CLI (Vercel)
npx skills add ulpi-io/skills --skill run-parallel-agents-feature-build -a claude-code2Manually: put the files in place
~/.claude/skills/run-parallel-agents-feature-build/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/run-parallel-agents-feature-build/
1Skills CLI (Vercel)
npx skills add ulpi-io/skills --skill run-parallel-agents-feature-build -a codex2Manually: put the files in place
.agents/skills/run-parallel-agents-feature-build/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add ulpi-io/skills --skill run-parallel-agents-feature-build -a cursor2Manually: put the files in place
.cursor/skills/run-parallel-agents-feature-build/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install ulpi-io/skills run-parallel-agents-feature-build2Manually: put the files in place
.github/skills/run-parallel-agents-feature-build/SKILL.md1Skills CLI (Vercel)
npx skills add ulpi-io/skills --skill run-parallel-agents-feature-build -a gemini-cli2Platform CLI
gemini skills install https://github.com/ulpi-io/skills --path run-parallel-agents-feature-buildManually: put the files in place
.devin/skills/run-parallel-agents-feature-build/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add ulpi-io/skills --skill run-parallel-agents-feature-build -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 run-parallel-agents-feature-build
How do I install run-parallel-agents-feature-build in Claude Code?
Run npx skills add ulpi-io/skills --skill run-parallel-agents-feature-build -a claude-code. The skill file is installed at ~/.claude/skills/run-parallel-agents-feature-build/SKILL.md.
What does run-parallel-agents-feature-build do?
It checks that build tasks are independent, assigns agents, and starts them concurrently in isolated worktrees. After they finish, it checks for conflicts and reports the outcome of each lane.
What work can this skill run in parallel?
It is for build tasks with at least three independent lanes, no shared write scope, and no dependencies between lanes. It is not meant for research, debugging, or tightly coupled work.