// Track

Building with Claude

The Anthropic API from First Call to Production

Build directly on the Claude API — the Messages API, the tool-use loop, structured outputs that survive the real validator, prompt caching economics, adaptive thinking, and the Batch API. Ends with a tool-using integration hardened by typed errors, a usage ledger, and a deploy-time smoke gate.

Recommended: Complete Prompt Engineering Mastery first

10 lessons~70 min total

Lessons are shown in recommended order. Complete them in sequence for the best experience — or jump to any lesson.

Lesson 566·6 min read

Meet the Claude API

One endpoint, three models, and the mental model that makes the rest of this track click.

Lesson 567·7 min read

Conversations, System Prompts & Stop Reasons

Statelessness is a design decision — your code owns history, and every stop reason is a signal your integration must handle.

Lesson 568·7 min read

Tool Use Part 1: The Agentic Loop

Claude asks, you execute, results return, repeat — the loop that powers every agent including Claude Code itself.

Lesson 569·7 min read

Tool Use Part 2: Reliable Tools

Quality engineering for tools: descriptions that fire, errors that feed back, and control modes that match your use case.

Lesson 570·6 min read

Structured Output That Survives Production

Green mocked tests cannot catch a production-illegal schema — the only proof is a live smoke call at deploy time.

Lesson 571·7 min read

Prompt Caching: The 90% Discount

Every turn re-sends full history — prompt caching is the mechanism that makes stateless-by-design economical at scale.

Lesson 572·6 min read

Adaptive Thinking & the Effort Dial

You don't know how much thinking a task needs — adaptive thinking lets Claude decide, and the effort dial sets the depth.

Lesson 573·9 min read

Streaming & the Batch API: Delivery Modes

Three delivery modes — sync, stream, batch — each optimal for a different latency-volume-cost profile; picking the wrong one is the most expensive mistake in Claude integrations.

Lesson 574·7 min read

Production Hardening: Errors, Retries & Cost Control

The gap between a Claude demo that works and a Claude integration that runs for months without incident is almost entirely error handling, typed exceptions, and cost visibility.

Lesson 575·8 min read

Capstone: Ship a Tool-Using Integration

Assemble every piece from this track into a production-grade support-ticket triage service — tool loop, schema output, prompt caching, typed-error retry wrapper, usage ledger, and a deploy-time smoke gate that proves it all works before you ship.