Does it support parallel execution?

Yes, it analyzes the tool's operations to design appropriate parallel architectures, such as work-stealing or map-reduce patterns, for the generated skill.

Can I use this for my own private APIs?

Yes, provided you supply the documentation or local README files, Skill Maker can analyze the specifications to generate a private, custom skill for your environment.

Why does it use SplitMix64 seeding?

SplitMix64 is used to provide deterministic behavior in the generated skills, ensuring that AI-driven tool interactions are predictable and reproducible across sessions.

How does Skill Maker discover tool documentation?

It uses Firecrawl to search the web and scrape API references, READMEs, and technical documentation to automatically extract operations, inputs, and code examples.

What is a Claude Code meta-skill?

A meta-skill is a specialized AI capability designed to create, manage, or improve other skills; Skill Maker specifically automates the generation of SKILL.md files for new tools.

Skill Maker

Name: Skill Maker
Author: plurigrid

byplurigrid

•

Developer Tools

Automatically generates production-ready Claude Code skills by analyzing tool documentation and applying standardized implementation patterns.

Skill Maker is a sophisticated meta-skill designed to accelerate the creation of specialized AI capabilities within the Claude Code ecosystem. By leveraging Firecrawl for documentation scraping and advanced semantic analysis, it transforms API references, READMEs, and technical specs into fully functional SKILL.md files. The factory follows a rigorous architecture including SplitMix64 deterministic seeding, ternary polarity mapping for output classification, and parallel execution design, ensuring that every generated skill is optimized for high-performance, predictable AI interactions.

Key Features

01Ternary output classification mapping (GF(3) = {-1, 0, +1})

02Template-driven SKILL.md generation for instant Claude Code integration

032 GitHub stars

04Deterministic seeding strategy generation using SplitMix64

05Automated tool discovery and documentation analysis using Firecrawl

06Parallel execution architecture design (Work-Stealing, Map-Reduce)

Use Cases

01Rapidly onboarding new CLI tools or APIs into the Claude Code environment

02Standardizing custom internal developer tools with consistent AI implementation patterns

03Creating high-performance parallelized skills for complex data processing workflows

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add plurigrid/asi skill-maker

For use in Claude.ai and ChatGPT

Download Skill