agent-tool-design
CommunityDesign robust tools for LLM agents
Software Engineering#mcp#error-handling#naming-conventions#idempotency#tool-design#input-schema#llm-agents
Authorabzhaw
Version1.0.0
Installs0
System Documentation
What problem does it solve?
Designing tools for LLM agents often suffers from inconsistent naming, unclear input schemas, and weak error handling, leading to brittle integrations across MCP bridges, cowork-mcp, and orchestrator layers.
Core Features & Use Cases
- One tool, one responsibility: keep actions isolated and easy to reason about.
- Descriptive names and rich descriptions: enable the LLM to select the correct tool with confidence.
- Explicit error semantics: always return useful errors and support safe retries.
- Idempotent by default: design tools that can be retried without side effects.
- MCP Tool Schema guidance: provide a clear template for input, outputs, and annotations to standardize tool definitions.
Quick Start
Define a new MCP tool by specifying its title, input schema, and idempotency, then implement and wire it to the backend.
Dependency Matrix
Required Modules
None requiredComponents
Standard package💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: agent-tool-design Download link: https://github.com/abzhaw/juliaz_agents/archive/main.zip#agent-tool-design Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.