Can this skill work with self-hosted deployments?

Yes, while the skill uses gimme.tools as the default domain, it is designed to be configurable for enterprise or self-hosted WebSpec environments.

What is the Three-Way Join algorithm in WebSpec?

It is a matching logic that ranks tools by combining three data sources: NLP intent extracted from the user request, the tool registry's embeddings, and the user's current authentication state (OAuth or Keychain).

What are the primary triggers for this skill?

This skill is triggered by phrases such as 'service discovery', 'nlp resolution', 'three-way join', 'tool matching', and 'embedding schema'.

How does this skill handle ambiguous user requests?

It uses multi-vector embeddings to find the closest semantic matches and can present multiple ranked options to the user if the intent similarity is close across different categories.

WebSpec Service Discovery

Name: WebSpec Service Discovery
Author: I-m-A-g-I-n-E

byI-m-A-g-I-n-E

0•

API開発

Optimizes tool matching and API resolution using NLP-driven intent extraction and a three-way join algorithm.

The WebSpec Service Discovery skill empowers Claude to intelligently bridge the gap between natural language requests and actionable API tools. By implementing a sophisticated three-way join algorithm—integrating user intent, service registries, and security credentials—it ensures that the most relevant and authorized tools are selected for any given task. This skill is essential for developers building agentic workflows where precise tool selection, semantic search, and seamless authentication handling are critical for reliable automation within the WebSpec ecosystem.

主な機能

01Context-aware UX ranking for Connected, Keychain, and Available services

02Three-way join algorithm combining registry data, OAuth status, and keychain hints

030 GitHub stars

04Weighted scoring system based on similarity, recency, and user preference

05Multi-vector embedding schema for robust semantic tool matching

06NLP-driven intent extraction with predicate-object normalization

ユースケース

01Developing natural language interfaces for complex enterprise ecosystems with multiple overlapping integrations

02Implementing secure tool discovery that prioritizes services the user has already authorized via OAuth

03Building AI agents that choose the correct API endpoint from a large registry based on vague user input

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add i-m-a-g-i-n-e/webspec service-discovery

For use in Claude.ai and ChatGPT

Download Skill