autoskill is an intelligent meta-skill that analyzes project context and problem statements to recommend and orchestrate the most relevant tools for any given task.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install autoskill
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 autoskill 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 autoskill?
autoskill acts as a sophisticated dispatcher for your AI development environment. Instead of manually searching for the right tool, you provide a problem description, and the skill performs a multi-phase analysis of your current git state, file extensions, and conversation history. It effectively bridges the gap between a user's intent and the vast library of available Openclaw Skills by scoring them based on relevance, domain, and technical stack.
Built with a safety-first architecture, it ensures that no action is taken without explicit user consent. It categorizes skills into recommended, suggested, or high-risk tiers, providing developers with a transparent execution plan. This makes it an essential tool for complex workflows where the optimal combination of specialized skills might not be immediately obvious.
autoskill Use Cases
- Automatically identifying the correct testing and linting tools after describing a bug.
- Routing to deployment and CI/CD skills while ensuring high-risk actions are gated behind manual approval.
- Analyzing a conversation to suggest refactoring or documentation skills without specific user prompts.
- Streamlining multi-step development tasks by sequencing relevant skills in order of priority.
- Discovering newly installed Openclaw Skills that match the current project's language signals.
How autoskill Works
- Problem Extraction: The skill inspects the local environment (git branch, language signals, and file changes) to build a structured context profile.
- Inventory Scan: It scans the active system registry to identify all available Openclaw Skills and groups them into functional buckets like testing, security, or backend.
- Relevance Scoring: Each candidate skill is scored (0-100) based on intent match, domain overlap, keyword relevance, and stack compatibility.
- Tier Classification: Skills are assigned to tiers: Recommended (score ≥ 70), Suggest (40-69), or High-Risk (based on security heuristics).
- Confirmation Gate: A mandatory execution plan is presented to the user, requiring explicit approval before any skill is invoked.
- Sequential Execution: Approved skills are executed one by one, with the outcome of each step logged in a final audit report.
autoskill Setup
To begin using this meta-routing capability, ensure it is available in your environment and invoke it via the CLI or your agent interface.
# Invoke with a specific problem description
/autoskill fix the auth middleware and add unit tests
# Invoke without arguments to analyze the current conversation context
/autoskill
autoskill Data Schema & Taxonomy
The skill generates a Context Profile and a Decision Audit Report to maintain transparency throughout the automation process.
| Data Component | Description |
|---|---|
| Context Profile | Includes Problem Statement, Action Intent, Stack (e.g., Python, TypeScript), and Domain Tags. |
| Scoring Table | A detailed breakdown of each skill's score, tier, and the specific reason for its inclusion. |
| Execution Plan | A preview list of skills queued for invocation, pending user confirmation. |
| Audit Report | A final summary showing which Openclaw Skills were applied, skipped, or blocked, including final outcomes. |
autoskill Advanced Features
- High-Risk Heuristics: Automatically flags skills involving payments, database migrations, or deployments as [HIGH-RISK], regardless of their relevance score.
- Language Signal Detection: Dynamically detects the project stack by inspecting files like package.json, Cargo.toml, or go.mod.
- Human-in-the-Loop Gate: Enforces a strict AskUserQuestion protocol, ensuring no automated actions occur without a manual confirmation signal.
- Intelligent Ranking: Limits recommendations to the top 5 highest-scoring skills to prevent runaway chaining and ensure focused task completion.
- Dynamic Registry Integration: Automatically recognizes and incorporates any newly added Openclaw Skills without requiring manual configuration updates.
SKILL.md
Loading
Related Openclaw Skills
Derivative Point Picking
A specialized mathematical reasoning skill designed to solve and prove complex derivative-based problems involving root existence, root counts, and parameter ranges.
New Job First 90 Days Plan
A strategic onboarding framework designed to help professionals navigate the first three months of a new role through structured learning, alignment, and execution.
Flight Delay Compensation Prep
An automated assistant designed to analyze flight disruptions and generate structured compensation claims based on global aviation regulations.
Circle K Stores Analysis Skill
A comprehensive research skill for analyzing the global operations, acquisition history, and business model of Circle K and its parent company, Alimentation Couche-Tard.
Chef Software Automation
An advanced configuration management skill that treats infrastructure as code using a Ruby-based DSL and a culinary-themed automation metaphor.
Chanel Paris
A specialized knowledge skill providing deep technical and business insights into the Chanel Paris luxury empire for AI-driven analysis.
Featured*
