Does this skill support the Model Context Protocol (MCP)?

Yes, it covers the emerging MCP standard, which is becoming the industry-standard language for connecting AI agents to external tools and data.

Why are tool descriptions more important than implementation?

The LLM never interacts with your source code; it only sees the tool's schema and description. A vague description causes the model to guess or hallucinate, regardless of how well the code is written.

How does this skill help reduce token costs?

By creating concise, unambiguous schemas and avoiding 'tool bloat,' you ensure the LLM understands the tool's purpose immediately without requiring long, repetitive context or multiple retry attempts.

What are the common anti-patterns in tool building?

Common mistakes include providing vague descriptions, allowing tools to fail silently without returning helpful error messages to the LLM, and overloading an agent with too many redundant tools.

Agent Tool Builder

Name: Agent Tool Builder
Author: claudiodearaujo

byclaudiodearaujo

•

API 개발

Architects high-performance AI agent interfaces through optimized JSON schemas, descriptive documentation, and robust error recovery patterns.

The Agent Tool Builder skill empowers developers to design reliable interfaces between Large Language Models and external systems. Recognizing that an LLM’s success depends entirely on the clarity of a tool's schema and description rather than its underlying code, this skill provides expert guidance on crafting unambiguous documentation, implementing the Model Context Protocol (MCP) standard, and building resilient error-handling mechanisms. It is essential for developers looking to eliminate agent hallucinations, prevent silent failures, and reduce token overhead in complex agentic workflows.

주요 기능

01Anti-pattern detection for tool bloat and silent failures

02Descriptive prompt engineering for function parameters

03Robust error handling and recovery strategies

04Model Context Protocol (MCP) integration patterns

051 GitHub stars

06JSON Schema optimization for LLM readability

사용 사례

01Refining function-calling schemas to reduce LLM hallucinations

02Optimizing tool descriptions to lower token consumption

03Building custom MCP servers for proprietary data sources

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro agent-tool-builder

For use in Claude.ai and ChatGPT

Download Skill