Subagent-Driven Development for Openclaw

A robust workflow for executing implementation plans by dispatching specialized subagents for coding and two-stage quality reviews.

user-wangjun
v1.0.1
Apr 7, 2026
1
741
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install subagent-driven-dev

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install subagent-driven-dev 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 Subagent-Driven Development?

Subagent-Driven Development is a specialized workflow within the Openclaw Skills ecosystem designed to streamline the execution of complex implementation plans. By isolating individual tasks and assigning them to fresh subagents, it prevents context pollution and ensures that every piece of code undergoes a rigorous two-stage review process focusing on specification compliance and technical quality. This approach allows developers to maintain high standards of code integrity while leveraging AI for rapid, automated iteration within a single session.

Subagent-Driven Development Use Cases

  • Executing pre-defined implementation plans with independent tasks.
  • Maintaining high code quality through automated, multi-stage AI reviews.
  • Rapidly iterating on features without human-in-the-loop bottlenecks between tasks.
  • Preventing context drift in long-running AI coding sessions using Openclaw Skills.

How Subagent-Driven Development Works

  1. Read the implementation plan and extract all tasks with full text and context into a Todo list.
  2. Dispatch a fresh implementer subagent for the specific task, providing complete context upfront to avoid file-reading overhead.
  3. The implementer subagent performs coding, testing, and self-review before committing the changes.
  4. A spec compliance reviewer subagent validates the work to ensure all requirements are met without over-building.
  5. A code quality reviewer subagent evaluates the technical implementation for best practices and approved patterns.
  6. The cycle repeats for all remaining tasks followed by a final global code review and branch finalization.

Subagent-Driven Development Setup

To get started with this strategy in Openclaw Skills, ensure your environment has the required prompt templates:

# Create or verify the required prompt templates in your skill directory
touch implementer-prompt.md
touch spec-reviewer-prompt.md
touch code-quality-reviewer-prompt.md

You must also have the following dependency skills active: superpowers:writing-plans and superpowers:finishing-a-development-branch.

Subagent-Driven Development Data Schema & Taxonomy

Component Description
implementer-prompt.md Context and instructions for the task-specific coding agent
spec-reviewer-prompt.md Rules for validating implementation against the original specification
code-quality-reviewer-prompt.md Standards for technical review and code health
TodoWrite The tracking mechanism for task completion status
Help Response Schema Structured XML format for handling subagent requests for help

Subagent-Driven Development Advanced Features

  • Automated two-stage review loops: Ensures both specification compliance and high-quality code through recursive refinement.
  • Context-isolated subagents: Prevents the implementation of one task from polluting the logic or context of another.
  • Structured help protocols: Standardized responses for ambiguity, blockers, conflicts, and scope decisions within Openclaw Skills.
  • TDD Enforcement: Naturally integrates Test-Driven Development into every subagent task lifecycle.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*