Domain 1
Agents and Workflows
Read the material that has been authored for this domain, alongside its published exam value.
- Exam weight
- 14.7%
- Approximate items
- About 8 items
Notes are not yet written
This domain has no authored note sections. The blueprint below shows what the domain measures while the notes are being contributed.
What this domain covers
| Sub-skill | Exam weight | Approximate items | What it measures | Status |
|---|---|---|---|---|
| Agent Construction with Claude | 5.3% | About 2.8 items | Claude Agent SDK; custom agent loops and harnesses; managed agent deployment models (self-hosted vs. Anthropic-hosted); hooks for deterministic actions | Scaffolded |
| Agent Patterns and Frameworks | 4.9% | About 2.6 items | Tool-use loops, sub-agents, memory, context-window management; agentic abstraction frameworks (Strands, LangGraph, PydanticAI) for multi-step tasks | Scaffolded |
| Agent Architecture | 4.5% | About 2.4 items | Decision criteria for workflow vs. agent; manager/supervisor hierarchies; the role of subagents in task execution | Scaffolded |
Reference application
The reference application does not yet demonstrate this domain.
Practice
This domain's self-check has two halves. Its recall prompts come from the notes and you grade yourself on them. Its quiz asks 8 items — this domain's share of a 53-item mock — and is scored by the same rule the mock uses. Only the scored quiz feeds a readiness signal.
The bank holds 8 of the 8 items the quiz asks for.