Chapters

The request/response contract every agent is built on — and the four fields the exam actually tests.

Descriptions are the selection mechanism. Schemas buy syntax, never semantics.

The agentic loop, hub-and-spoke, and why hooks beat prompts for anything that costs money.

One flat tool namespace, config that decides who gets the server, and errors an agent can act on.

Where a config file lives decides who it reaches — plus plan mode, CI with -p, and session forking.

Examples instead of adjectives, criteria instead of judgement, and retries that carry the actual error.

Half price, a 24-hour ceiling that is not an SLA, and no tool loops. Plus surgical recovery via custom_id.

Fixed pipeline when the scope is knowable, dynamic when it emerges — and why 14 files need 15 passes.

Rule-based triggers, the three that only look like triggers, and a handoff the operator can read cold.

Categories that decide retryability, the four ways information gets destroyed, and gaps you annotate instead of hide.

Facts blocks that survive compaction, trimming at the boundary, and delegation as a context strategy.

Claims that carry their source and date, conflicts kept rather than resolved, and format chosen by content.

One job per tool, Grep-first investigation, and what to do when Edit has nothing unique to match.