From product-design
Executes multiple Claude Code agents in parallel via cpo CLI: validates manifests, creates git worktrees, runs dependency-based waves, monitors progress. Use for parallel tasks and orchestration.
How this skill is triggered — by the user, by Claude, or both
Slash command
/product-design:parallel-executionThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Execute multiple Claude Code agents in parallel using the `cpo` (Claude Parallel Orchestrator) CLI.
Execute multiple Claude Code agents in parallel using the cpo (Claude Parallel Orchestrator) CLI.
The cpo tool handles all execution complexity: git worktrees, wave dependencies, and progress monitoring.
pip install claude-parallel-orchestrator
# or
pipx install claude-parallel-orchestrator
| Command | Description |
|---|---|
cpo validate <dir> | Validate manifest structure and prompts |
cpo run <dir> | Execute all waves (respects dependencies) |
cpo status <dir> | Check execution status |
# 1. Validate before execution
cpo validate parallel/TS-0042-inventory-system/
# 2. Execute parallel agents
cpo run parallel/TS-0042-inventory-system/
# 3. Monitor progress (in another terminal)
cpo status parallel/TS-0042-inventory-system/
cpo run Doeslogs/ and report.jsonTasks execute in waves based on dependencies:
Wave 1: task-001, task-002, task-003 (parallel - no deps)
↓ wait for completion
Wave 2: task-004, task-005 (parallel - depend on Wave 1)
↓ wait for completion
Wave 3: task-006 (sequential - depend on Wave 2)
Each wave waits for all tasks in the previous wave to complete before starting.
Agents run with --dangerously-skip-permissions because they're isolated in worktrees:
For programmatic orchestration in CI/CD or custom workflows:
// orchestrator.ts
import { ClaudeAgent } from '@anthropic-ai/claude-agent-sdk';
import { readdir, readFile } from 'fs/promises';
import { join } from 'path';
async function runParallelTasks(parallelDir: string) {
const tasksDir = join(parallelDir, 'tasks');
const contextFile = join(parallelDir, 'context.md');
const context = await readFile(contextFile, 'utf-8');
const tasks = await readdir(tasksDir);
const agents = tasks
.filter(f => f.endsWith('.md'))
.map(async (taskFile) => {
const taskPath = join(tasksDir, taskFile);
const taskContent = await readFile(taskPath, 'utf-8');
const agent = new ClaudeAgent({
systemPrompt: `You are implementing a task.
Context: ${context}
Follow contracts in ${parallelDir}/contracts/.`,
});
return agent.run(`Execute this task:\n\n${taskContent}`);
});
const results = await Promise.all(agents);
return results;
}
runParallelTasks('parallel/TS-0042-inventory-system');
Agents signal completion by creating a marker file:
touch .claude-task-complete
This enables:
cpo to detect task completion| Method | Best For | Parallelism |
|---|---|---|
| cpo CLI | Standard workflow, most users | True parallel with wave deps |
| Claude Code SDK | CI/CD, custom orchestration | Fully programmable |
cpo validate before cpo runlogs/task-*.log for debugging/parallel-integrate after all tasks completeAfter execution:
parallel/TS-0042-inventory-system/
logs/
task-001.log # Agent output
task-002.log
...
report.json # Execution summary
integration-report.md # Generated by /parallel-integrate
parallel-agents skill: Overall workflow and directory structureparallel-decompose skill: Creating tasks before executionparallel-prompt-generator skill: Generate prompts from task specsagent-tools skill: Tool permissions (for granular control)/parallel-integrate command: Post-execution verificationnpx claudepluginhub jpoutrin/product-forge --plugin product-designOrchestrates parallel Claude agent development tasks using git worktrees and cpo CLI. Validates directories, checks status, and executes waves from decomposed parallel folders.
Launches parallel subagents via the Task tool with run_in_background: true to concurrently implement features, analyze directories, or run independent tasks.
Dispatches independent tasks to concurrent Claude Code agents, checks for file and dependency conflicts, and integrates results. Useful when 3+ truly independent tasks exist across different subsystems.