claude-api

You can integrate the Claude API by installing the anthropic Python package or @anthropic-ai/sdk for TypeScript, then calling messages.create to send prompts. This Skill provides patterns for streaming, vision, tools, and extended thinking across both SDKs.

Yes, the Claude Agent SDK supports agent workflows with multi-step reasoning. You can build automated tasks, chatbots, and copilots by combining the SDK with prompt caching and tool integration patterns covered by this Skill.

Yes, this Skill covers streaming responses and prompt caching using the Anthropic SDK. It provides best practices for handling streaming data, optimizing costs, and managing error handling within your Python or TypeScript applications.

The best way to handle tool use and vision capabilities is through the Claude Messages API patterns provided here. It demonstrates how to pass images, define tools, and structure responses for applications like code assistants and automated workflows.

You need the anthropic Python package or @anthropic-ai/sdk for TypeScript to use extended thinking and batches. This Skill outlines the required SDK setup and model selection guidance to implement these advanced features effectively.

Model selection and cost optimization are needed to manage API expenses and performance when building Claude applications. This Skill offers guidance on choosing appropriate models and applying prompt caching to reduce costs during real-world tasks.