A comprehensive AI-driven code review agent that analyzes PR diffs for structural integrity, data safety, and design consistency while proactively applying auto-fixes.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install gstack-review
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install gstack-review using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The Pre-Landing PR Review is a high-performance automation for Openclaw Skills designed to sit between your code and the merge button. Unlike standard linters, this skill performs a deep, two-pass analysis of your git diffs, focusing on critical areas such as SQL safety, race conditions, and LLM trust boundaries. It ensures that every PR is not only technically sound but also aligned with the original project intent and design standards.
By leveraging a Fix-First heuristic, this skill reduces the friction of code reviews by automatically repairing mechanical issues and batching high-level design judgments for user approval. It acts as a tireless technical lead, cross-referencing your changes against documentation, TODO lists, and even external adversarial review tools like Codex to ensure your codebase remains robust and debt-free.
To integrate this review capability into your Openclaw Skills environment, ensure you have the GitHub CLI installed and then configure the skill checklist:
# Ensure GitHub CLI is authenticated
gh auth status
# Create the mandatory checklist directory
mkdir -p .claude/skills/review/
# Initialize your project-specific checklist
touch .claude/skills/review/checklist.md
You can also optionally add a design-checklist.md in the same directory for frontend-specific audits.
The skill utilizes a structured feedback loop and specific file lookups to maintain context:
| Data Component | Source / Location | Purpose |
|---|---|---|
| Checklist | .claude/skills/review/checklist.md |
Core rules for the two-pass review process. |
| Design Specs | DESIGN.md or design-system.md |
Reference for UI/UX consistency checks. |
| Task Context | TODOS.md |
Used to detect scope drift and update task status. |
| Branch Metadata | gh pr view |
Fetches base branch and original PR intent. |
| Status Codes | DONE, BLOCKED, etc. |
Standardized protocol for reporting review completion. |
Loading
A sophisticated AI agent skill that transforms text outlines into high-impact presentation decks featuring a consistent IP character and professional YouMind-style illustrations.

A comprehensive toolkit for writing, debugging, and porting GLSL shaders across WebGL, Three.js, and React Three Fiber runtimes.

A comprehensive integration for managing PagerDuty incidents and on-call schedules via the REST API.

A high-stakes brainstorming and diagnostic tool that transforms raw ideas into rigorous technical design documents using YC-inspired methodology.

A specialized AI agent skill designed to provide deep technical and fundamental analysis of Chinese A-shares, Hong Kong stocks, and US-listed Chinese companies.

A professional-grade content generation tool designed to create high-impact scripts, titles, and publishing packages specifically for WeChat Channels (Channels IP).








































