tooluniverse-custom-tool

Register and call private REST APIs or local Python tools in ToolUniverse.

Updated Apr 18, 2026
One-click install
npx skills add https://github.com/Centaurioun/osteogenesis_imperfecta --skill tooluniverse-custom-tool
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tooluniverse-custom-tool
Source: https://github.com/Centaurioun/osteogenesis_imperfecta/tree/main/.agents/skills/tooluniverse-custom-tool
Command: npx skills add https://github.com/Centaurioun/osteogenesis_imperfecta --skill tooluniverse-custom-tool

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Enables developers and workspace maintainers to add, validate, and run private or custom tools alongside ToolUniverse's built-in toolset so internal APIs, offline calculators, and reusable plugins can be used securely and consistently.

Core Features & Use Cases

  • Workspace JSON tools: Drop JSON configs into a .tooluniverse/tools directory to register REST-backed tools without writing Python.
  • Python class tools: Implement local-only tools with @register_tool classes for custom logic and deterministic computation.
  • Plugin packages: Package reusable tools as pip-installable plugins with entry points and discoverable JSON configs.
  • Validation & deployment: Use tu test, return_schema, and profile.yaml to verify tool behavior, load tools into MCP/tu serve, and manage workspace discovery.

Quick Start

Create a .tooluniverse/tools JSON or Python tool, add test_examples, then run tu test MyAPI_search to validate it.

Frequently Asked Questions about tooluniverse-custom-tool

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

FAQPage Schema
How do I add a custom REST API tool to ToolUniverse using JSON config?

To add a custom REST API tool, drop a JSON config file into the .tooluniverse/tools directory to register REST-backed tools without writing Python code. You define parameters and return_schema validation directly in the JSON config.

How do I create local Python class tools for deterministic computation in ToolUniverse?

You can implement local-only Python class tools by using the @register_tool decorator pattern. This allows you to run custom logic and deterministic computation securely within your workspace environment.

What is the best way to validate custom tool behavior before deploying to MCP server?

The best way to validate custom tool behavior is using the tu test CLI command. You can run tu test MyAPI_search with test_examples to verify JSON parameter validation and return_schema before loading tools into MCP or tu serve.

Can I package reusable custom tools as pip-installable plugins?

Yes, you can package reusable custom tools as pip-installable plugins with entry points and discoverable JSON configs. This allows workspace maintainers to share internal APIs and offline calculators consistently across projects.

Does ToolUniverse support profile-based workspace discovery for custom tools?

Yes, ToolUniverse supports profile-based workspace discovery using profile.yaml. This mechanism manages how local Python tools and JSON configs are discovered and loaded into the MCP server or tu serve environments.