A high-performance orchestration workflow that dispatches multiple AI agents in parallel to handle coding, testing, and review phases using isolated git worktrees.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-team-pipeline
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agent-team-pipeline using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Agent Team Pipeline?
The Agent Team Pipeline is a sophisticated orchestration framework designed to maximize the efficiency of Openclaw Skills. It functions by transforming OpenClaw into a central command hub that dispatches specialized Codex instances to handle distinct roles in the development process. By assigning specific agents to coding, testing, and reviewing tasks simultaneously, developers can replicate a full-scale engineering team's output within a single automated environment.
This skill is particularly effective for projects that require rigorous quality assurance and peer review. By utilizing isolated git worktrees, the pipeline ensures that each agent operates in a sandbox environment, preventing file conflicts and maintaining a clean version history while multiple Openclaw Skills processes run in tandem.
Agent Team Pipeline Use Cases
- Implementing complex features that require dedicated, separate phases for logic implementation, test coverage, and peer review.
- Executing independent development tasks in parallel to significantly reduce the total time-to-ship.
- Creating a structured, multi-agent development environment for mission-critical codebases that demand high compliance and validation.
How Agent Team Pipeline Works
- The user provides a high-level development objective to OpenClaw, which acts as the lead architect.
- OpenClaw dispatches specific sub-tasks to the Coder, Tester, and Reviewer agents via specialized task distribution scripts.
- Each agent operates within its own dedicated git worktree, allowing for parallel execution without overlapping file changes.
- The Coder generates the feature logic, while the Tester concurrently develops test suites and the Reviewer analyzes code quality.
- A feedback loop is established where the Coder addresses Reviewer comments until the code meets all defined quality gates.
Agent Team Pipeline Setup
To deploy this skill within your Openclaw Skills environment, first initialize the isolated git worktrees for your team roles:
cd ~/.openclaw/workspace
git worktree add -b coder ./agent-coder HEAD
git worktree add -b tester ./agent-tester HEAD
git worktree add -b reviewer ./agent-reviewer HEAD
Next, create a task distribution script to route commands to the appropriate directory:
#!/bin/bash
# Example distribution logic
case "$1" in
coder) cd "./agent-coder" && codex exec "$2" ;;
tester) cd "./agent-tester" && codex exec "$2" ;;
reviewer) cd "./agent-reviewer" && codex exec "$2" ;;
esac
Agent Team Pipeline Data Schema & Taxonomy
The Agent Team Pipeline manages data through a distributed directory structure to ensure isolation between Openclaw Skills agents:
| Role | Directory Path | Primary Data Output |
|---|---|---|
| Coder | agent-coder/ |
Feature source code and implementation files |
| Tester | agent-tester/ |
Test scripts, unit tests, and validation suites |
| Reviewer | agent-reviewer/ |
Review logs, feedback summaries, and compliance checks |
All metadata regarding the synchronization of these agents is handled via the underlying git worktree metadata and branch references.
Agent Team Pipeline Advanced Features
- Parallel execution support for simultaneous Coder, Tester, and Reviewer workflows.
- Git worktree isolation to prevent merge conflicts during concurrent AI operations.
- Automated feedback loops that trigger iterative code improvements based on Reviewer critiques.
- Custom role extensibility allowing users to add specialized agents (e.g., Security Auditor) to the pipeline.
- Integration with Openclaw Skills task distribution for seamless command-line control over the entire agent team.
SKILL.md
Loading
Related Openclaw Skills
Data Governance Skill
A comprehensive toolset for managing data quality, metadata, and compliance across multiple database environments.
write-paper
An AI agent skill that automates the drafting of complete academic research papers in LaTeX format by analyzing project code and experimental data.
A-Stock Real-Time Analysis & Volume Distribution
A comprehensive stock analysis tool for China A-Shares providing real-time pricing, minute-level volume breakdown, and portfolio profit/loss tracking.
Grafana Readonly
A secure, read-only interface for exploring Grafana dashboards, extracting panel queries, and rerunning analytics without risking configuration changes.
Influencer & Trendy Store Guide
A specialized skill for discovering, filtering, and planning visits to viral influencer stores and niche local boutiques.
ocas-triage
A deterministic system scheduler and priority queue manager designed to govern task execution and mentor handoffs for AI agents.
Featured*
