Parallel Agent Workflow
Coordinate multiple agents working in parallel using git worktrees to avoid file conflicts. Use when needing concurrent agents to modify different files/components with worktree isolation, or when asked to use worktrees for parallel agent work.
Install
npx skills add bolasblack/claude-skills --skill parallel-agent-workflowgit clone --depth 1 https://github.com/bolasblack/claude-skills.git /tmp/claude-skills
mkdir -p ~/.claude/skills
cp -r /tmp/claude-skills/skills/parallel-agent-workflow ~/.claude/skills/parallel-agent-workflowPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Parallel Agent Workflow
Execute multi-agent refactoring, feature development, or parallel work using git worktrees to prevent conflicts.
## How It Works
1. **Split work** across N agents, each in isolated git worktree
2. **Agents work independently** on different components/files
3. **Merge in stages** - combine related work into logical groups
4. **Final merge** - all work combined into single branch
## Instructions
### Phase 1: Planning
1. **Identify the work units** - What needs to be done?
2. **Determine agent count**:
- If user specifies: Use that count
- If not specified: Auto-determine based on independent work units
3. **Ask for missing requirements** if not in environment:
- Setup script (if `dev-setup` doesn't exist): "What command sets up dependencies?"
- Test command (if unknown): "What command verifies the build?"
4. **Create merge strategy**:
- Group related work units logically
- Plan merge stages (N agents → M groups → 1 final)
- Example: 6 agents → 2 groups → 1 final
### Phase 2: Execute Agent Work
For each agent:
1. **Create worktree**:
```bash
# Pattern: {project}@{branch-name}
git worktree add ../{project}@{work-unit-name} -b {branch-name}
```
- **CRITICAL**: If worktree creation fails → STOP IMMEDIATELY, report error
2. **Agent instructions must include**:
- cd into worktree
- Run setup script (or `dev-setup` if exists)
- Run test command BEFORE changes
- Do the workDetails
- Skill name
parallel-agent-workflow- Repository
- bolasblack/claude-skills
- Path
skills/parallel-agent-workflow/SKILL.md- Repo updated
- 18 November 2026
FAQ
What does the Parallel Agent Workflow skill do?
Coordinate multiple agents working in parallel using git worktrees to avoid file conflicts. Use when needing concurrent agents to modify different files/components with worktree isolation, or when asked to use worktrees for parallel agent work.
How do I install the Parallel Agent Workflow skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add bolasblack/claude-skills --skill parallel-agent-workflow. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Parallel Agent Workflow popular?
It lives in bolasblack/claude-skills, which has 5 GitHub stars and 0 forks.
Related skills
Caveman
5bolasblack/claude-skills
Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".
Agent Centric
5bolasblack/claude-skills
Track decisions (AGD) with validation and indexing. Use when making design decisions, recording important choices, discussing trade-offs, or when user mentions AGD or decision record.
Codex
5bolasblack/claude-skills
OpenAI Codex second opinion via the codex CLI — three modes. Review: independent diff review with a pass/fail gate. Challenge: adversarial pass that tries to break the code. Consult: ask Codex anything, with session continuity for follow-ups. Use when the user says "codex review", "codex challenge", "ask codex", or wants a second opinion from Codex.
Process Level Event Storming
5vnovakovits/claude-skills
Apply Alberto Brandolini's Process Modeling level of Event Storming — modeling ONE business process in full sticky-note grammar (read model → actor → command → aggregate/external system → event → policy → …) on a timeline, with hot spots as first-class output. Covers the picture-that-explains-everything loop, grammar rules, completeness walks (backwards walk, orphan checks, missing failure paths), swimlane/frame layout for remote boards (Miro/Mural/FigJam), and mapping the model onto process managers, sagas, and workflow code. Use when designing or documenting a single process end-to-end, mode
Orchestration
5catch-the-wave/fullstack-ios-claude-skills
Coordinate supervised Orca workers: threaded messages, blocking ask/reply, task dispatch, worker_done/escalation waits, task DAGs, decision gates, coordinator loops, and decomposing work across agents. Use `orca-cli` for full ownership handoffs — "hand off", "handoff", "handover", "give this to another agent", "another worktree" — unless asked to supervise, monitor, or coordinate a DAG, and for terminal control, lightweight terminal prompts, shell commands, Orca worktree management, and reading or waiting on terminals.
Project Orchestrator
8Freddy-CHEN-ux/00_claude_skills
Orchestrate multi-phase project implementation using background Claude agents for isolated development, review, and integration.
Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.