A high-precision execution framework that dispatches fresh subagents for independent tasks with automated two-stage review cycles.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install subagent-development
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 subagent-development 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 methodology designed to execute implementation plans by isolating tasks into fresh subagent sessions. This approach prevents context pollution and ensures that each unit of work is handled with maximum focus. As a premier entry in the ecosystem of Openclaw Skills, it bridges the gap between high-level planning and low-level execution by maintaining a single-session workflow while leveraging the power of multiple specialized agents.
The skill is built on the core principle of two-stage validation: every task must first pass a specification compliance review to ensure requirements are met, followed by a rigorous code quality review. This ensures that the final output is not only functional but also maintains professional architectural standards without requiring constant human intervention between tasks.
Subagent-Driven Development Use Cases
- Executing detailed implementation plans where tasks are independent and can be handled sequentially.
- Maintaining high code quality and TDD standards in long-running development sessions.
- Projects where strict adherence to technical specifications is mandatory to avoid over-engineering or missing requirements.
- Situations where context bloat in a single AI session leads to performance degradation or hallucination.
How Subagent-Driven Development Works
- The controller reads the implementation plan and extracts all tasks with full text, context, and requirements.
- A TodoWrite object is created to track the lifecycle and completion status of every task in the plan.
- A fresh implementer subagent is dispatched for the current task to write code, run tests, and perform a self-review.
- A spec reviewer subagent is invoked to verify that the implementation matches the original requirements exactly.
- Once spec-compliant, a code quality reviewer subagent evaluates the implementation for architectural integrity and clean code standards.
- The subagent iterates on any issues found during reviews until both stages are approved, at which point the task is marked complete.
Subagent-Driven Development Setup
To install this skill and integrate it into your workflow, run the following command in your terminal. This will add the necessary templates to your Openclaw Skills environment:
npx clawhub@latest install subagent-development
Subagent-Driven Development Data Schema & Taxonomy
The skill organizes its workflow through specialized Markdown prompt templates and state tracking:
| Component | File / Method | Description |
|---|---|---|
| Implementer | implementer-prompt.md |
The primary agent responsible for coding, testing, and committing work. |
| Spec Reviewer | spec-reviewer-prompt.md |
Specialized subagent that validates code against the implementation plan. |
| Quality Reviewer | code-quality-reviewer-prompt.md |
Specialized subagent focused on linting, architecture, and best practices. |
| Task Tracker | TodoWrite | An internal data structure used to manage the execution queue and status. |
Subagent-Driven Development Advanced Features
- Context-curated dispatching which provides subagents with exactly the information they need, reducing token waste and overhead.
- Multi-stage quality gates that strictly enforce spec compliance before allow code quality reviews to begin.
- Automated re-review loops where implementers must fix issues and resubmit to the same reviewer for final approval.
- Seamless integration with other Openclaw Skills like plan-writing and development branch finalization for a complete end-to-end SDLC.
SKILL.md
Loading
Related Openclaw Skills
Startup Metrics
A comprehensive framework for tracking and optimizing core startup KPIs from Seed through Series A.
shadcn-ui
A comprehensive skill for building accessible, high-performance user interfaces using shadcn/ui, Radix UI, and Tailwind CSS.
Local Testnet Setup for Uniswap
A comprehensive tool to launch a local Anvil testnet forking live chains with pre-deployed Uniswap contracts and multi-token funded accounts.
Setup DCA for Uniswap
A comprehensive automation skill for establishing non-custodial dollar-cost averaging strategies on Uniswap with optimized routing and safety guards.
Subagent-Driven Development
A robust framework for executing complex implementation plans by dispatching fresh AI subagents for every task with a mandatory two-stage review process.
Submit UniswapX Limit Order
A specialized skill for executing gasless, MEV-protected limit orders via UniswapX Dutch auctions.
Featured*
