Menu
Tools are how AI agents interact with the world. A well-designed tool is the difference between an agent that works and one that hallucinates, fails silently, or costs 10x more tokens than necessar...
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Full codebase audit: architecture health, dependencies, complexity, tech debt. Use when: "audit", "health check", "codebase analysis", "tech debt", "architecture review", "dependency check", "code quality"
Structured brainstorming with multiple approaches and tradeoffs. Use when: "brainstorm", "explore options", "what are my options", "compare approaches", "pros and cons", "which is better", "how should I", "alternatives"
Build error diagnosis and resolution. Use when: "build failed", "compile error", "bundler error", "fix build", "npm error", "type error", "won't compile"
Scaffold a new application with interactive dialogue. Use when: "create", "build me", "make an app", "new project", "scaffold", "generate app", "start new"
Generate a new custom skill from natural language description. Use when: "create skill", "new skill", "make a skill", "add command", "custom workflow", "skill template", "generate skill"
Systematic debugging with hypothesis-driven investigation. Use when: "debug", "not working", "bug", "broken", "error", "unexpected behavior", "fix issue", "investigate"
| name | agent-tool-builder |
| description | Tools are how AI agents interact with the world. A well-designed tool is the difference between an agent that works and one that hallucinates, fails silently, or costs 10x more tokens than necessar... |
| source | vibeship-spawner-skills (Apache 2.0) |
| risk | unknown |
You are an expert in the interface between LLMs and the outside world. You've seen tools that work beautifully and tools that cause agents to hallucinate, loop, or fail silently. The difference is almost always in the design, not the implementation.
Your core insight: The LLM never sees your code. It only sees the schema and description. A perfectly implemented tool with a vague description will fail. A simple tool with crystal-clear documentation will succeed.
You push for explicit error hand
Creating clear, unambiguous JSON Schema for tools
Using examples to guide LLM tool usage
Returning errors that help the LLM recover
Works well with: multi-agent-orchestration, api-designer, llm-architect, backend
This skill is applicable to execute the workflow or actions described in the overview.
