All Articles
Diagnosing Claude API Prompt Cache Misses — How to Read the usage Field
If your Claude API prompt cache isn't reducing your bill, the usage field is where to start. This guide walks through the five most common reasons cache_read_input_tokens stays at zero and how to fix each one.
Managing Claude API Prompts as Code: Registry, Versioning, and A/B Testing in Production
Anyone running Claude API in production eventually hits the same wall: which prompt was served, when, to whom, and at what version? This guide walks through a registry-based architecture with A/B testing, gradual rollouts, and automatic rollback — all implementable yourself in TypeScript.
Production Infrastructure for Claude API — 8 Things You Need Between 'It Works' and 'It Holds Up'
There is a much bigger gap than you'd think between a working Claude API call on your laptop and a service that survives real users. Here are the eight pieces of infrastructure I now consider non-negotiable, learned the hard way.
Claude API "Spanner Temporarily Unavailable" — A Field Guide to Handling It in Production
The first time the Claude API returned 'Spanner temporarily unavailable', I genuinely could not tell if the bug was on Anthropic's side or mine. This guide unpacks what the error actually means, the three-tier retry strategy I now run in production, and the metrics I added to my monitoring after three real incidents.
Indie Developer's Claude API SaaS Launch Blueprint — A 90-Day Roadmap from Idea to Paying Customers
A complete 90-day roadmap for building an indie Claude API business: idea validation, Stripe integration, SEO, subscription pricing tests, and the operational and emotional discipline that makes it last. Drawing on twelve years of solo app development and the new realities of AI APIs.
Claude API Revenue Models Compared — Pay-per-Use, Subscription, One-Shot, and Revenue Share for Indie Developers
When you build something with Claude API to make money, the hardest decision is rarely the API itself — it's the pricing model. This guide compares the four most common models from an indie developer's point of view.
Production-Grade Hallucination Defense for Claude API: A Multi-Layer Architecture
Prompt engineering alone is not enough to suppress hallucinations in production. After a real customer incident, I rebuilt the system around four defensive layers — input grounding, tool-use escape hatches, citations, and post-hoc verification. This is the implementation playbook.
Replay-Driven Testing for Claude API: A Production Pattern for Recording and Replaying Responses
A production-grade design for stabilizing Claude API tests by recording and replaying real responses. Covers cassettes for Messages, Streaming, Tool Use, CI integration, and incident replay.
Fixing Claude's 'tool result could not be submitted' Error — A Practical Recovery Guide
When Claude shows the dreaded 'tool result could not be submitted' banner, the conversation gets stuck and even fresh prompts fail. This is the field-tested recovery guide I built after one rough night with my own MCP server.
Building a Scalable Real-Time AI Chat Server with Claude API × WebSocket × Redis Pub/Sub — Node.js Production Architecture, Multi-User Management, and Cost Control
Production implementation of a real-time AI chat server using Claude API, WebSocket, and Redis Pub/Sub. Covers SSE vs WebSocket trade-offs, scalable Node.js connection management, JWT auth, and per-user cost control.
Decoding Claude's 'Spanner Temporarily Unavailable' Error and How to Handle It
The 'Spanner temporarily unavailable' error occasionally appears in Claude API and Claude.ai responses. This guide unpacks what the message reveals about Anthropic's infrastructure and walks through practical retry strategies that production teams actually use.
How I Cut My Claude API Bill in Half With Prompt Caching
Done right, Anthropic's prompt caching can roughly halve your monthly API spend on workloads with long, repeated system prompts. Here is the design playbook I use after six months of running it in production.