claude-api

Guide Claude API integration across languages and implementation patterns.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/stellarone/evaluate --skill claude-api-stellarone
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: claude-api
Source: https://github.com/stellarone/evaluate/tree/main/.agents/skills/claude-api
Command: npx skills add https://github.com/stellarone/evaluate --skill claude-api-stellarone

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Claude integration guidance can be confusing because you must choose the correct API surface (single call, workflow, tool use, or Agent SDK) and map it to your project language and needs.

Core Features & Use Cases

  • Surface selection with guardrails: Recommends the simplest Claude interface tier that matches your task complexity and whether you need tools/agents/file/web access.
  • Language-aware documentation routing: Detects the project language (Python, TypeScript, Java, Go, Ruby, PHP, C#, or cURL) and directs you to the relevant Claude API examples.
  • Implementation-ready best practices: Covers streaming, tool use (tool runner vs manual loop), structured outputs, error handling, batching, files, and compaction considerations.

Quick Start

Ask an AI to “Show me how to call the Claude API for a TypeScript app with tool use, including streaming and a safe manual loop for tool results.”

Frequently Asked Questions about claude-api

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

FAQPage Schema
How do I choose the right Claude API surface for my application workflow?

To implement Claude API tool use, you can utilize the SDK tool runner or build a safe manual agentic loop. This Skill provides implementation patterns for correctly handling tool results and managing agentic loops across supported languages.

Can I use the Claude API with TypeScript and streaming?

When handling Claude API errors, apply robust error-handling guidance to manage failures safely. This Skill ensures correct API usage requirements for managing model selection constraints, compaction behavior, and tool or result handling.

What is the difference between single-message requests and multi-step workflows in the Anthropic SDK?

The difference between single-message requests and multi-step workflows is the complexity of interaction. Single-message requests handle one-off prompts, while multi-step workflows manage extended interactions requiring state tracking and tool use orchestration.

Does the Claude API support structured outputs and file access?

Yes, the Claude API supports structured outputs and file access. This Skill guides you through the correct implementation patterns for utilizing these features, including batching and compaction considerations across supported languages and raw HTTP.