Agents and Workflows Questions
Practice questions for Agents and Workflows topic in Claude Certified Developer – Foundations. 27 questions covering this domain.
A document analysis agent retrieves records from a customer database. In testing, retrieved records contain text such as "Ignore your instructions and...
A developer is building a production agent that can send emails and delete files. Before executing irreversible actions, which architectural pattern s...
An agent using Claude Haiku 4.5 with extended thinking is running a multi-hour session. The developer notices that after adding a non-tool-result user...
A developer builds a long-horizon agentic task where a faster executor model handles routine steps but occasionally needs strategic guidance from a mo...
An agent uses Context editing (beta) during a long tool-calling session. After 30 tool calls, the context is nearly full due to accumulated tool resul...
What is the key characteristic of Claude Managed Agents (Sessions API) that distinguishes it from standard stateless Messages API calls?
A production customer service agent occasionally enters infinite loops — repeatedly calling the same tool with similar inputs for dozens of iterations...
What does the term "agentic loop" refer to in Claude integrations?
What are the three main API families that make up Claude Managed Agents?
A developer uses the Anthropic Python SDK and wants to avoid writing the manual tool execution loop where they parse tool_use blocks, execute tools, a...
A developer built an agent using the Claude Sessions API. The project is now complete, and the developer wants to free the stored session transcript f...
In a multi-agent system, a worker agent receives tool results from a third-party data provider. The results contain embedded text such as "System over...
A developer builds an agent that calls a database lookup tool followed by a calculation tool where the calculation input depends on the database resul...
An enterprise agent handles wire transfer requests. Risk policy requires all transfers over $50,000 to be approved by a senior manager before executio...
A compliance-critical financial agent executes wire transfers. The system prompt instructs Claude to refuse transfers exceeding $100,000 without manag...
In an agentic tool-calling loop, how does the application know that Claude has finished the task and is providing the final answer rather than calling...
A document processing agent calls multiple tools in sequence: extract_text → analyze_sentiment → generate_summary. During production use, the agent fr...
An agent retrieves customer data from a database and uses it to answer questions. Occasionally, a database connection timeout causes the database look...
A developer builds a customer service agent that escalates complex issues to human operators. The escalation logic is currently based on Claude's self...
A production agent uses 8 different tools for customer support. During testing, Claude occasionally calls the wrong tool (e.g., calling create_ticket ...
Sign in to see all 27 questions
Create a free account to browse the questions included with the free plan.