claude-api

Build Claude-powered apps and agents using the API and Agent SDK.

Updated Apr 3, 2026
One-click install
npx skills add https://github.com/BEIRUX/agent-skills --skill claude-api-beirux
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: claude-api
Source: https://github.com/BEIRUX/agent-skills/tree/main/claude-api
Command: npx skills add https://github.com/BEIRUX/agent-skills --skill claude-api-beirux

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides concrete, opinionated guidance and language-specific examples for building applications and agents that use the Claude API and Anthropic Agent SDK, removing guesswork around model selection, streaming, tool use, and integration patterns.

Core Features & Use Cases

  • Language detection and curated walkthroughs for Python, TypeScript, Java, Go, Ruby, C#, and PHP to jumpstart integrations.
  • Best-practice defaults (use Claude Opus 4.6, adaptive thinking, streaming) and detailed recipes for streaming, prompt caching, compaction, structured outputs, batching, and file uploads.
  • Tool-use and agent patterns including the beta tool runner, manual agentic loop, server-side code execution, web search/fetch, memory tool, and MCP integration for building autonomous assistants.

Quick Start

Show me a Claude API example that streams a response and handles tool calls in the project's primary language.

Frequently Asked Questions about claude-api

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

FAQPage Schema
How do I stream responses and handle tool calls with the Claude API in Python?

To stream Claude API responses with tool calls in Python, use language-specific SDK examples that provide streaming patterns, tool-runner integration, and best-practice defaults like Claude Opus 4.6 with adaptive thinking enabled.

Can I use the Anthropic Agent SDK with Java, Go, Ruby, or PHP?

Yes, the Anthropic Agent SDK supports Java, Go, Ruby, C#, and PHP. It provides curated walkthroughs and language-specific SDK examples to jumpstart integrations for building chat UIs, agents, and structured outputs across these languages.

What is the best way to implement prompt caching and compaction workflows for Claude agents?

The best way to implement prompt caching and compaction workflows is by following detailed recipes that cover streaming, prompt caching guidance, and compaction patterns, ensuring efficient context management for autonomous Claude-powered agents.

Does the Claude API support structured outputs and batch processing?

Yes, the Claude API supports structured outputs and batch processing. It provides specific examples and recipes for batching requests, file uploads, and generating structured data formats within your application integration.

How do I integrate MCP and web search tools into a Claude agent loop?

To integrate MCP and web search tools into a Claude agent loop, use the provided tool-use and agent patterns including the beta tool runner, manual agentic loop, server-side code execution, web search/fetch, and MCP integration.

What model should I default to when building Claude-powered applications?

You should default to Claude Opus 4.6 when building Claude-powered applications. This model offers best-practice defaults including adaptive thinking and streaming, ensuring optimal performance for agents and chat interfaces.