Pass the Claude
Code cert.
The exam is multiple-choice over six known scenarios. The path to a clean pass is the same every time: drill the scenarios, hit 95% on simulated mocks, then book it.
Reps gives you the question bank, the mock exam, and the per-domain dashboard that tells you exactly where you're weak.
No signup. No payment. Drill, see your gaps, fix them.
The six scenarios
Anthropic's exam guide ships five scenarios; we cover all six known deep-dives. Memorize them — every passer says they were the bulk of the prep.
Programmatic enforcement, structured handoffs, escalation calibration.
CLAUDE.md hierarchy, slash commands, plan vs direct, iterative refinement.
Hub-and-spoke orchestration, parallel spawning, provenance preservation.
Built-in tool selection, MCP scoping, session management, context degradation.
Non-interactive CLI, structured output, multi-pass review, batch vs sync.
tool_use + JSON schema, validation/retry, batch correlation, provenance.
Six ways to drill
Quick warm-up to full mock exam. Pick the mode that matches where you are this morning.
Mock exam
50 questions, 120-minute timer, weighted by domain like the real thing. Hit 95% before you book the test.
Scenario drills
Lock onto one of the six scenarios. Repeat until pattern-matching is automatic.
Domain drills
Target your weakest domain (D1–D5). The dashboard tells you which one to fix next.
Gotcha drills
The anti-patterns the exam slips in as distractors: sentiment-based escalation, CLAUDE_HEADLESS, batch-for-pre-merge, bigger-model-fixes-dilution.
Socratic tutor
When MC drilling isn't enough, the AI tutor walks you through the underlying concept and pushes back on shallow answers.
Read the scenarios
All six exam-guide deep-dives, in-app, formatted for re-reading. Every passer says: memorize them.
Tips from someone who passed
Notes from a recent passer. We baked these into how Reps drills.
Drill till you pass.
720 is the pass line. 950 is when you stop drilling and book the exam. Reps tells you which line you're on.
Start your first drill