Notes from the prep
Short pieces on what's working as we build out the curriculum and exam strategy.
- 2026-05-12
Updating your study plan for the May 2026 model lineup
With Opus 4.7, Sonnet 4.6, and Haiku 4.5 all current, here's what to revisit in our curriculum this week.
- 2026-05-10
Choosing between Opus 4.7, Sonnet 4.6, and Haiku 4.5
A decision flowchart for the current model lineup. Memorize this — at least one exam question is a near-verbatim variant.
- 2026-05-09
Haiku 4.5 for production: where it actually wins
Haiku 4.5 punches above its price tier. Three patterns where swapping it in saves real money without quality loss.
- 2026-05-08
Sonnet 4.6: the workhorse most candidates should default to
Sonnet 4.6 is the model the exam assumes when scenarios don't specify. Know its sweet spot and its trade-offs cold.
- 2026-05-07
Claude Opus 4.7 with 1M context: what changes for the exam
Anthropic released Opus 4.7 with a 1M-token context window. The exam objectives don't change — but how you answer some scenarios will.
- 2026-05-06
Shipping your first agent to production
A pragmatic checklist: evals, observability, budgets, kill switch.
- 2026-04-30
Three recipes for bulletproof structured output
JSON mode alone isn't enough. Combine it with examples, validation, and a single retry.
- 2026-04-23
Skills vs. slash commands: when to use which
Slash commands are for things you invoke; skills are for things you teach.
- 2026-04-16
Leave headroom: the 70% rule
Plan for ~70% of the window so you have room for outputs and unexpected tool result blocks.
- 2026-04-09
Using Claude Code hooks for team-wide policy
Hooks are the right place to enforce 'no force-push' and 'no skipping tests' across every developer's machine.
- 2026-04-02
Why you should build agents without tools first
Before adding tools, see if a pure-LLM loop solves your task. Most teams over-engineer.
- 2026-03-25
Mock exam strategy: how to use your 60 minutes
Two passes. First pass: 90 seconds per question, mark anything uncertain. Second pass: spend the rest on the marked ones.
- 2026-03-19
Prompt cache economics: why static-first ordering matters
The 5-minute cache cuts input cost by ~90% on hits. Almost everyone loses cache hits by putting volatile content first.
- 2026-03-11
MCP in 15 minutes
The Model Context Protocol in plain language: tools, resources, and prompts — when to use which.
- 2026-03-04
What to study first for the Claude Certified Architect exam
Start with agentic patterns and tool design — they're the two biggest scoring blocks and the most transferable to real work.
- 2026-02-26
The 80/20 of the certification
Twelve concepts get you to 720. Memorize these and the rest is gravy.
- 2026-02-12
Setting up your prep environment
Get an Anthropic API key, a sandbox repo, and Claude Code wired up before you read anything else. You'll absorb 2-3x more.
- 2026-02-05
Reading Anthropic's docs efficiently
Three reading orders that compress weeks of reading into a long weekend.
- 2026-01-29
The exam blueprint: what 'foundations' means in practice
Five domains, 27/18/20/20/15 weighting. 'Foundations' doesn't mean trivia — it means production patterns most teams get wrong on day one.
- 2026-01-15
Why we built this site
The official Claude Certified Architect exam launched. We needed a single-page path through 5 sprawling domains. So we built one.