Anthropic’s Claude Certified Architect Exam (CCA-F): The Agent That Did Exactly What the Injected…

Anthropic's Claude Certified Architect Exam (CCA-F) Part 7 argues that AI agent safety is an architecture problem, not a prompting problem, and that real safety lives at the tool boundary with a deny, ask, or allow gate. The article covers five ideas for hardening agents: grounding against hallucination, prompt-injection defense with a deterministic gate, a hardcoded-over-softcoded trust hierarchy, human-in-the-loop approval gates, and red-team-evals-fallback discipline.

Member-only story Anthropic’s Claude Certified Architect Exam CCA-F : The Agent That Did Exactly What the Injected Text Told It To CCA-F Part 7: Why AI agent safety is an architecture problem, not a prompting problem, and how that single shift in thinking is the judgment a Claude certification scenario actually tests You can spend a week word-smithing the perfect “please refuse malicious instructions” system prompt and still watch one line of untrusted text walk your agent into a destructive action. Real safety lives at the tool boundary, in a deny, ask, or allow gate that ranks the operator over the user over anything the model merely reads. In this article:You will learn why the safest answer to an AI agent safety question is almost never “improve the prompt.” We cover the five ideas that turn a fragile agent into a hardened one: grounding against confident hallucination, prompt-injection defense with a deterministic gate, the hardcoded-over-softcoded trust hierarchy, human-in-the-loop approval gates, and the red-team-evals-fallback discipline that measures quality…