Superclaw is a comprehensive workflow engine that prevents AI agents from coding prematurely by enforcing a strict design, plan, and execute cycle.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install superclaw
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 superclaw 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 Superclaw?
Superclaw is a disciplined software development framework designed specifically for the Openclaw Skills ecosystem. Based on the original superpowers framework, it introduces a rigorous methodology that forces agents to prioritize architectural integrity over immediate code generation. By chaining three distinct phases—Brainstorming, Writing Plans, and Executing Plans—it ensures that every line of code is backed by a peer-reviewed design and a granular implementation strategy.
This skill package is essential for developers who want to avoid the common pitfalls of AI coding, such as compounding errors and lack of context. It integrates directly with local memory files and workspace conventions to provide a transparent, auditable, and highly reliable development experience. Whether you are building complex CLI tools or modular web components, Superclaw provides the guardrails necessary for professional-grade software engineering.
Superclaw Use Cases
- Creating new software features or standalone CLI tools where design trade-offs must be evaluated.
- Building complex components that require a step-by-step implementation plan to avoid logic errors.
- Adding functionality to existing codebases while maintaining strict documentation and memory logs.
- Implementing Test-Driven Development (TDD) workflows where the plan must precede the code.
How Superclaw Works
- Brainstorming Phase: The agent analyzes context from MEMORY.md and USER.md, asks Socratic questions to define requirements, and proposes 2-3 approaches. No code is written until a design document is approved and saved.
- Writing Plans Phase: Once a design is approved, the agent asks about preferred methodology (like TDD) and breaks the work into 2-5 minute tasks, saving a structured implementation plan.
- Executing Plans Phase: The agent loads the plan and batches tasks into groups of maximum 5. It uses sessions_spawn for isolation and requires a manual checkpoint review before moving to the next batch.
- Progress Tracking: Throughout the lifecycle, the agent updates daily logs and memory to ensure the workflow is resumable and transparent.
Superclaw Setup
To install this skill package within your environment, run the following command in your terminal:
npx clawhub@latest install superclaw
The components of this package will automatically load when the agent detects a task related to software creation or feature implementation.
Superclaw Data Schema & Taxonomy
Superclaw organizes all design and implementation artifacts within the workspace to ensure persistent documentation. The following directory structure and naming conventions are used:
| Directory | File Pattern | Description |
|---|---|---|
workspace/docs/plans/ |
YYYY-MM-DD-<topic>-design.md |
Contains approved architectural designs and trade-off analysis. |
workspace/docs/plans/ |
YYYY-MM-DD-<topic>-plan.md |
Contains the granular task list and implementation methodology. |
memory/ |
YYYY-MM-DD.md |
Daily logs updated after every execution batch checkpoint. |
root |
MEMORY.md / USER.md |
Referenced for global project context and user preferences. |
Superclaw Advanced Features
- Socratic Questioning: Prevents impulsive coding by forcing a dialogue regarding constraints and trade-offs.
- Automatic Skill Chaining: Seamlessly transitions from design to planning to execution without manual intervention.
- Isolated Execution: Leverages sessions_spawn to create fresh sub-agents for each task, preventing environment pollution.
- Batched Checkpoints: Enforces a hard gate after every 5 tasks, requiring user confirmation to prevent error cascading.
- Methodology Flexibility: Supports multiple development styles including TDD and direct implementation based on user input.
SKILL.md
Loading
Related Openclaw Skills
YouTube Video Uploader
Securely upload videos to YouTube using the official Data API v3 and OAuth 2.0 authentication.
YouTube Upload
A technical skill for automating video uploads to YouTube via official OAuth 2.0 authentication and API integration.
Xiaohongshu Automation (xiaohongshu-auto)
An automated solution for managing Xiaohongshu accounts, publishing notes, and tracking performance data via Openclaw Skills.
git-workflow
A comprehensive automation skill for Git that handles change detection, commit message generation, and remote synchronization.
Investing Analyst
A thesis-driven trading assistant for equity and options analysis designed to enforce discipline and edge identification.
DexScreener
A CLI-driven integration for fetching real-time decentralized exchange market data, liquidity metrics, and token profiles.
Featured*
