CCDV-F Agents and the Agent SDK Practice Test 2 — 15 Questions

Question 1

When an agent is asked to perform a complex task, which approach minimizes latency while ensuring accuracy?

Accepted Answer

Decomposing the task into smaller, chained agent actions.. Breaking down a complex task into smaller, sequential tool-use steps allows the model to reason incrementally. This approach reduces the load per turn, makes errors easier to isolate, and improves overall accuracy by allowing the model to process feedback at each stage. While it increases the total number of turns, it is often faster than forcing a single, massive inference that may fail entirely.

Answer

Requesting the model to solve the entire problem in one turn.

Answer

Using the maximum possible context window for every request.

Answer

Adding 'think step-by-step' to the final response request.

Question 2

Which THREE strategies are effective for improving the reliability of an agent in production environments? (Select exactly 3)

Accepted Answer

Implementing strict input and output schema validation.. Production-grade agents require proactive management. Robust schema validation catches input errors early; comprehensive monitoring provides visibility into agent performance and failure modes; and human-in-the-loop patterns for high-stakes tasks provide a critical safety net. Together, these practices form a defense-in-depth strategy that ensures the agent performs reliably under real-world conditions where unpredictability is the norm.

Answer

Increasing the model's temperature to 1.5 for variety.

Answer

Hardcoding every possible user query path.

Question 3

Which of the following is a best practice when providing feedback to an agent after a failed tool call?

Accepted Answer

Provide the specific error message returned by the tool.. Providing specific, actionable error messages is essential for recovery. An agent cannot fix a failed tool call if it does not understand why it failed. By feeding back the exact error message (e.g., 'invalid argument', 'permission denied'), you allow the model to adjust its future attempts, enabling the agent to learn from its errors and eventually succeed in the task without manual intervention.

Answer

Ignore the error and let the agent retry with the same input.

Answer

Restart the entire conversation history from scratch.

Answer

Send a generic 'try again' message to the agent.

Question 4

You are designing a multi-agent system where a 'Router' agent directs tasks to a 'Research' agent or a 'Writer' agent. What is the most efficient way to maintain state when the Research agent completes its task and needs to hand back control?

Accepted Answer

Appending the Research agent's tool results to the main message thread.. Handoffs in multi-agent systems rely on passing the conversation history or a summarized state between specialized instances. This ensures that the Router has the necessary context to decide the next step without losing the progress made by the Research agent. Proper state management prevents redundant work and keeps the agentic flow aligned with the user goal.

Answer

Persisting the Research agent's output to a global database for the Router.

Answer

Restarting the entire session with the Research output as a new prompt.

Answer

Using a separate API key for each agent to isolate their environments.

Question 5

When defining a tool for an agent using the Anthropic SDK, which field is required to ensure the model understands the specific schema of the input arguments?

Accepted Answer

The 'input_schema' property using JSON Schema format.. The input_schema field is mandatory because it uses JSON Schema to define the structure, types, and constraints of the data the agent must provide. Without this schema, the model cannot reliably generate valid tool calls that match the underlying function's expectations, leading to runtime errors and failed executions in the agentic workflow.

Answer

The 'examples' array within the tool definition.

Answer

A 'system_prompt' specific to that individual tool.

Answer

The 'output_format' property for return values.

Question 6

Refer to the exhibit. A developer is configuring an MCP server to provide an agent with database access. What is the primary purpose of the 'args' field in this configuration block?

Accepted Answer

It specifies the command-line flags and parameters used to initialize the server.. The 'args' field provides the necessary runtime parameters for the MCP server process. In this specific exhibit, it tells the 'uvx' command which package to run and which specific database file to mount. Correct configuration of these arguments is vital for the agent to successfully connect to the intended external resource.

Answer

It defines the JSON Schema for the tools provided by the SQLite server.

Answer

It lists the specific SQL queries the agent is permitted to execute.

Answer

It acts as a whitelist for authorized users of the SQLite server.

Question 7

An agent is performing a 'Computer Use' task that requires taking multiple screenshots and moving the cursor. The model generates a tool_use block for 'screenshot' and 'mouse_move' in a single turn. How should the SDK-based orchestrator handle this response to ensure reliable execution?

Accepted Answer

Process the tools sequentially and return all results in the next turn.. Reliable computer use requires strict sequential execution of actions because each step modifies the visual state of the environment. If the orchestrator attempts to parallelize these actions, the 'screenshot' might capture the screen before the 'mouse_move' is finished, leading to incorrect state feedback for the model and causing the agent to fail its task.

Answer

Execute both tools in parallel to minimize latency for the user.

Answer

Wait for the model to issue a separate turn for each individual tool.

Answer

Only execute the first tool and ignore all subsequent blocks.

Question 8

A developer is implementing a 'Human-in-the-loop' (HITL) pattern for an agent that performs financial transactions. Which TWO strategies are most important for maintaining the security and reliability of this agentic workflow?

Accepted Answer

Providing the human with a clear diff of the proposed transaction data.. HITL is essential for high-stakes tasks where model errors could have significant real-world consequences. By introducing a manual approval step, developers can ensure that the agent remains within its defined boundaries. Clear communication of the agent's intent and a robust mechanism for human rejection are key pillars of this safety architecture.

Answer

Allowing the agent to bypass human approval if its confidence score is high.

Answer

Obfuscating the tool names from the human to prevent cognitive overload.

Answer

Using a shorter system prompt to ensure the agent acts more autonomously.

Question 9

An agent using the SDK encounters an error where it generates a tool call for a tool that does not exist in its configuration. Which adjustment to the agent's setup is most likely to resolve this hallucination?

Accepted Answer

Refining the system prompt to clearly define the boundaries of available tools.. Hallucinating non-existent tools usually occurs when the model is aware of a capability but hasn't been given the structured definition for it, or when the system prompt is ambiguous. Refining the system prompt to explicitly list the available tools and their purposes helps ground the model's reasoning in the provided technical context.

Answer

Enabling 'force_tool_use' to ensure it only uses one of the existing tools.

Answer

Switching the model from Claude 3.5 Sonnet to Claude 3 Haiku.

Answer

Increasing the max_tokens parameter to allow for more reasoning.

Question 10

When implementing a 'Computer Use' agent, which THREE capabilities are provided by the standardized Anthropic beta tools 'computer', 'text_editor', and 'bash'?

Accepted Answer

Capturing screenshots and performing mouse/keyboard interactions.. Anthropic provides specialized tools for computer interaction to ensure consistent behavior across different environments. These tools allow the agent to perform GUI actions, modify files with precision, and execute shell commands. Together, they form a comprehensive suite for agents that need to operate as a virtual developer or assistant on a computer system.

Answer

Rewriting entire binary files to optimize system performance.

Answer

Automatically upgrading the host operating system to the latest version.

Question 11

Refer to the exhibit. An agent receives this response after attempting to use a 'fetch_url' tool. How will this specific block affect the agent's next reasoning step?

Accepted Answer

The model will interpret this as a signal to retry or handle the failure.. The 'is_error' flag is a critical signal to the model that the requested action did not succeed. By explicitly marking the result as an error, the SDK helps Claude understand that it needs to perform error recovery—such as retrying the request, checking the URL for typos, or reporting the failure to the user—rather than proceeding as if the data was retrieved.

Answer

The model will ignore the result and repeat the exact same tool call.

Answer

The model will treat the error message as the successful content of the URL.

Answer

The SDK will automatically terminate the session to prevent further errors.

Question 12

In the context of the Anthropic Agent SDK, what is the primary role of the 'Orchestrator' component?

Accepted Answer

Managing the loop between model reasoning and tool execution.. The Orchestrator acts as the central brain of the agentic loop, managing the cycle of sending prompts to the model, receiving tool calls, executing those tools, and feeding the results back. It abstracts the complexity of message management and ensures that the conversation flows logically between the AI and the available external capabilities.

Answer

Generating the final CSS styles for the agent's user interface.

Answer

Storing the model's weights on the local file system for offline use.

Answer

Compressing images before they are sent to the vision-capable model.

Question 13

Which TWO of the following are essential components of a 'tool_use' block generated by Claude?

Accepted Answer

A unique 'id' string used to match results to the specific request.. A 'tool_use' block must contain specific identifiers so that the SDK knows which tool to run and how to track its result. The 'id' uniquely identifies the specific call in the conversation history, while the 'name' maps the call to the actual function or service defined in the agent's configuration.

Answer

The full source code of the function being called.

Answer

The 'timestamp' of when the model decided to use the tool.

Answer

The 'confidence_score' the model assigns to this specific tool call.

Question 14

An agent is designed to manage a user's calendar. During a session, the user asks to 'Schedule a meeting for tomorrow at 2 PM.' The agent must first check for conflicts and then create the event. How should the developer handle the 'observation' phase in the SDK to ensure the agent doesn't double-book?

Accepted Answer

Feed the 'check_conflicts' tool_result back to the model before it issues 'create_event'.. The observation phase involves feeding the actual results of a tool (the calendar check) back into the agent's context. If the agent doesn't receive the output of the conflict check before it attempts to create the event, it is 'flying blind.' Ensuring the orchestrator waits for the first tool's result before proceeding is key to logical consistency.

Answer

Use a single tool that both checks for conflicts and creates the event.

Answer

Parallelize both calls to improve the speed of the calendar update.

Answer

Prompt the model to assume there are no conflicts to save tokens.

Question 15

When deploying an agent that uses the Model Context Protocol (MCP), what is the primary security risk of allowing the agent to dynamically discover and connect to any local MCP server?

Accepted Answer

The agent could gain unauthorized access to local files or system resources.. Dynamic discovery without strict authorization can expose sensitive local data or system capabilities to the agent. If an agent connects to a malicious or unintended MCP server, it could be tricked into executing harmful commands or leaking private information. Implementing a whitelist of trusted servers is a fundamental security practice for agentic systems.

Answer

The agent might use too many tokens while browsing the server's tool list.

Answer

The model might become confused by having too many tools available.

Answer

The MCP protocol might slow down the model's inference speed.