Manage AI tool and function calling with provider-agnostic definitions.
npx skills add https://github.com/arunoda/OkBrain --skill tools-arunoda
Please help me install this Agent Skill. Skill: Tools Source: https://github.com/arunoda/OkBrain/tree/main/.agent/skills/tools Command: npx skills add https://github.com/arunoda/OkBrain --skill tools-arunoda
This Skill provides a robust architecture for AI tool and function calling, enabling AI models to interact with external services and local data.
search_places tool to find nearby restaurants, then use compute_routes to get directions, and finally use get_weather_by_location to check the weather for the trip.Use the tools skill to search for restaurants near the Golden Gate Bridge.
High-intent search queries and answers about installing and using this skill.
You can define AI function calling tools using a provider-agnostic definition format that works with both Gemini and OpenAI. This allows you to specify the tool structure once and execute it across multiple AI providers without changing the core definitions.
The system includes pre-built tools for Google Maps (places, routes, weather, air quality), internet search via Brave and Tavily, news search, URL content reading, and calendar event management to facilitate immediate API orchestration.
You can add custom tools by defining their specific structure and providing an executor function. This extensibility allows the AI model to interact with external services and local data through your newly defined execution logic.
Yes, an AI assistant can orchestrate multiple tools in sequence, such as using a search tool to find nearby restaurants, a routing tool to compute directions, and a weather tool to check conditions for the trip.
The best way to manage external API integrations is using a provider-agnostic tool definition architecture. This enables AI models to interact with external services and local data through a unified executor function and structured tool definitions.
vm0-aiGenerate AI videos from images, text, or existing videos via the Runway API.
veniceaiProcess user inputs via Venice's Responses API into typed reasoning, messages, tool calls, and web searches.
Aqsagull99Detect user intents from natural language and map them to Todo operations.
juanreSwitch between LLM providers and access raw SDKs via get_provider().