tool-design

Design agent tool specifications with clear triggers, inputs, outputs, and error recovery.

1|Updated May 16, 2026
One-click install
npx skills add https://github.com/enigmaicon-eng/AI-Enterprise-OS --skill tool-design-enigmaicon-eng
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tool-design
Source: https://github.com/enigmaicon-eng/AI-Enterprise-OS/tree/main/examples/agent-skills/skills/tool-design
Command: npx skills add https://github.com/enigmaicon-eng/AI-Enterprise-OS --skill tool-design-enigmaicon-eng

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Tool-design helps you write agent-facing tool descriptions that remove ambiguity, so language models can reliably select the right tool and produce correctly formatted calls.

Core Features & Use Cases

  • Tool descriptions as contracts: Turns tool usage requirements into an unambiguous, agent-inferable interface.
  • Consolidation and architectural reduction: Reduces overlapping or redundant tools so agents don’t confuse similar capabilities.
  • Error recovery-ready design: Specifies actionable error conditions and recovery guidance to help agents self-correct.
  • Consistent schemas and conventions: Enforces stable parameter naming, return formats, and section structure to prevent description rot.

Use cases include designing new agent tools, debugging tool selection failures, standardizing tool interfaces across a codebase, and improving reliability when integrating MCP (Model Context Protocol) tools.

Quick Start

Use the tool-design skill when you are creating or revising agent tool descriptions so the model can choose the correct tool and call it with the right parameters.

Frequently Asked Questions about tool-design

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write agent tool descriptions that prevent incorrect tool selection?

Prevent agent tool selection failures by refactoring tool specifications into unambiguous contracts with clear usage triggers, precise inputs and outputs, and consistent schema conventions to reduce overlap-induced selection errors.

Why does my language model call the wrong tool or pass invalid parameters?

Language models call the wrong tool or pass invalid parameters when tool descriptions lack clear usage triggers and precise input and output contracts. Structuring tool descriptions as unambiguous agent-facing interfaces resolves this ambiguity.

What is the best way to consolidate overlapping agent tools?

Consolidate overlapping agent tools through architectural reduction, applying consistent schemas and naming conventions to reduce redundant capabilities and prevent description rot that causes selection confusion.

How do I design error recovery guidance for API contracts in agent tools?

Design error recovery guidance for API contracts by specifying actionable error conditions and structured recovery steps directly within the tool description to help agents self-correct during execution.

Does this approach work for standardizing MCP tool naming and routing?

Yes, this approach works for MCP integration by applying consistent naming schemas and routing conventions to tool descriptions, ensuring reliable tool selection and correctly formatted calls within the Model Context Protocol.

When should I refactor an agent toolset to fix tool selection failures?

Refactor an agent toolset to fix tool selection failures when overlapping capabilities cause description rot, requiring architectural reduction and stable parameter naming to restore reliable model inference.