> ## Documentation Index > Fetch the complete documentation index at: https://opentools.page/docs/llms.txt > Use this file to discover all available pages before exploring further. # Introduction > Build AI-first apps with ready-made tools OpenTools is a provider-agnostic Python SDK for building AI-first applications with reliable, production-ready tooling. It gives you ready-made sets of tools grouped by domain returning consistent outputs so that you can swap providers, frameworks, or models without rewriting your core logic. ## Start here Install OpenTools and run your first tool in minutes. ## How it works OpenTools is built around three simple ideas: * **Modules** are domains like *Trading*. * **Providers** are implementations like *Coinbase*. * **Adapters** connect OpenTools to agent frameworks like *LangGraph* or models like *Qwen*. Tools return **defined schemas**, so your app stays predictable even when the underlying provider changes. ## Core concepts What a tool is in OpenTools, and how tool specs are defined. Consistent outputs so you don’t rewrite logic per provider. ## Modules Modules group tools under one domain and define what is normalised. **Trading** is our first. See the tools, providers, and compatibility matrix. ## Adapters Adapters let you use OpenTools inside your agent stack without writing glue code. Use OpenTools inside agentic frameworks like PydanticAI or LangGraph. Use LLM providers from OpenAI to OpenRouter. ## Resources What’s next: modules, providers, and runtime improvements. Support expectations, security notes, and project policies.