Topic 05 Practice Test 1 covers Agentic AI, Tool Calling, Human Review, Strands, Lambda, ECS, and MCP in Domain 2 (Task 2.1). For broader exam preparation, review the AIP-C01 Exam Dumps. Every option includes focused technical reasoning tied to the tested decision.
Question 1
A support agent must remember the customer’s selected account and unresolved steps across several tool calls in the same case. Which approach should the developer use? Choose ONE.
- MCP for agent-tool interoperability
- Specialist-model routing
- Strands Agents with explicit session state
- Lambda action implementation
Correct Answer: C
Correct Answer
Answer C is correct because Strands Agents with explicit session state — in this scenario, uses the Strands agent framework, in this case, while preserving the state needed, in this case, across turns or workflow steps. Useful when an agent must retain task/session, at this layer, context while coordinating tools. Need: preserve conversational task state across tool calls. Strands Agents with explicit session state matches it directly.
Incorrect Answers
Answer A is incorrect because MCP for agent-tool interoperability — in this scenario, uses the Model Context Protocol, in this case, to expose tools through a, in this case, consistent agent-to-tool interface. Useful when tools should be reachable through, at this layer, a standardized protocol instead of, at this layer, bespoke integrations. Need: preserve conversational task state across tool calls. MCP for agent-tool interoperability solves another problem.
Answer B is incorrect because Specialist-model routing — in this scenario, routes a request to a, in this case, model selected for the task’s, in this case, capability, latency, or quality needs. Useful when different task classes benefit from, at this layer, different model strengths. Need: preserve conversational task state across tool calls. Specialist-model routing solves another problem.
Answer D is incorrect because Lambda action implementation — in this scenario, runs bounded custom business logic, in this case, for an agent tool or, in this case, action group in a serverless, in this case, function. Useful when a tool action is lightweight,, at this layer, stateless, and fits Lambda execution, at this layer, constraints. Need: preserve conversational task state across tool calls. Lambda action implementation solves another problem.
Question 2
A research workflow has separate agents for legal review, financial analysis, and technical validation, then combines their contributions. What should the team implement? Choose ONE.
- Strands Agents with explicit session state
- RETURN_CONTROL action group
- Lambda action implementation
- AWS Agent Squad multi-agent coordination
Correct Answer: D
Correct Answer
Answer D is correct because AWS Agent Squad multi-agent coordination — in this scenario, coordinates multiple specialized agents so, in this case, each can contribute a bounded, in this case, capability to a larger task. Useful when a workflow benefits from several, at this layer, specialist agents rather than one, at this layer, monolithic agent. Need: coordinate several specialist agents. AWS Agent Squad multi-agent coordination matches it directly.
Incorrect Answers
Answer A is incorrect because Strands Agents with explicit session state — for this design, uses the Strands agent framework, for this workload, while preserving the state needed, for this workload, across turns or workflow steps. Fits when an agent must retain task/session, for the stated need, context while coordinating tools. Constraint: coordinate several specialist agents. Strands Agents with explicit session state targets another layer.
Answer B is incorrect because RETURN_CONTROL action group — in this scenario, returns the predicted action and, in this case, elicited parameters to the calling, in this case, application instead of invoking Lambda, in this case, directly. Useful when the application itself must decide, at this layer, how or where to execute, at this layer, the requested action. Need: coordinate several specialist agents. RETURN_CONTROL action group solves another problem.
Answer C is incorrect because Lambda action implementation — for this design, runs bounded custom business logic, for this workload, for an agent tool or, for this workload, action group in a serverless, for this workload, function. Fits when a tool action is lightweight,, for the stated need, stateless, and fits Lambda execution, for the stated need, constraints. Constraint: coordinate several specialist agents. Lambda action implementation targets another layer.
Question 3
A platform team wants the same agent client to call tools hosted by multiple teams without writing a custom protocol for each tool. Which option most directly addresses the requirement? Choose ONE.
- API Gateway feedback endpoint
- Specialist-model routing
- Standardized function schema
- MCP for agent-tool interoperability
Correct Answer: D
Correct Answer
Answer D is correct because MCP for agent-tool interoperability — for this design, uses the Model Context Protocol, for this workload, to expose tools through a, for this workload, consistent agent-to-tool interface. Fits when tools should be reachable through, for the stated need, a standardized protocol instead of, for the stated need, bespoke integrations. Constraint: standardize agent-to-tool interoperability across teams. MCP for agent-tool interoperability is the closest fit.
Incorrect Answers
Answer A is incorrect because API Gateway feedback endpoint — in this scenario, provides an HTTP interface through, in this case, which applications can submit structured, in this case, human feedback or approval data. Useful when the workflow needs an external, at this layer, feedback collection endpoint. Need: standardize agent-to-tool interoperability across teams. API Gateway feedback endpoint solves another problem.
Answer B is incorrect because Specialist-model routing — for this design, routes a request to a, for this workload, model selected for the task’s, for this workload, capability, latency, or quality needs. Fits when different task classes benefit from, for the stated need, different model strengths. Constraint: standardize agent-to-tool interoperability across teams. Specialist-model routing targets another layer.
Answer C is incorrect because Standardized function schema — in this scenario, defines tool names, parameters, types,, in this case, and required fields consistently so, in this case, the model can call tools, in this case, reliably. Useful when tool invocation errors are caused, at this layer, by ambiguous or inconsistent function, at this layer, definitions. Need: standardize agent-to-tool interoperability across teams. Standardized function schema solves another problem.
Question 4
An autonomous remediation workflow must branch, retry selected steps, and stop after a bounded number of failed attempts. Which design best satisfies this requirement? Choose ONE.
- Custom result aggregation
- Lambda timeout guard
- Step Functions orchestration
- MCP client library
Correct Answer: C
Correct Answer
Answer C is correct because Step Functions orchestration — in this scenario, models multi-step workflow control, branching,, in this case, retries, approvals, and stopping conditions, in this case, explicitly. Useful when an AI workflow needs deterministic, at this layer, orchestration around model or tool, at this layer, calls. Need: make multi-step control flow and retries explicit. Step Functions orchestration matches it directly.
Incorrect Answers
Answer A is incorrect because Custom result aggregation — in this scenario, combines outputs from multiple model, in this case, or agent workers using explicit, in this case, application logic. Useful when several parallel specialist results must, at this layer, be reconciled into one answer. Need: make multi-step control flow and retries explicit. Custom result aggregation solves another problem.
Answer B is incorrect because Lambda timeout guard — in this scenario, enforces bounded execution around a, in this case, tool or model-supporting function so, in this case, stalled work cannot run indefinitely. Useful when a lightweight serverless action needs, at this layer, a clear execution timeout. Need: make multi-step control flow and retries explicit. Lambda timeout guard solves another problem.
Answer D is incorrect because MCP client library — in this scenario, gives agents a consistent client-side, in this case, mechanism for discovering and invoking, in this case, MCP-exposed tools. Useful when multiple MCP servers should be, at this layer, accessed through a common client, at this layer, pattern. Need: make multi-step control flow and retries explicit. MCP client library solves another problem.
Question 5
A lightweight tool can occasionally hang on a downstream socket and must never run longer than the approved execution window. What is the most appropriate implementation choice? Choose ONE.
- Lambda timeout guard
- Circuit breaker around failing tool
- Explicit stopping condition
- Step Functions retry with backoff
Correct Answer: A
Correct Answer
Answer A is correct because Lambda timeout guard — for this design, enforces bounded execution around a, for this workload, tool or model-supporting function so, for this workload, stalled work cannot run indefinitely. Fits when a lightweight serverless action needs, for the stated need, a clear execution timeout. Constraint: bound one serverless tool execution. Lambda timeout guard is the closest fit.
Incorrect Answers
Answer B is incorrect because Circuit breaker around failing tool — in this scenario, temporarily stops repeated calls to, in this case, a failing dependency after an, in this case, error threshold and allows controlled, in this case, recovery. Useful when a repeatedly failing downstream tool, at this layer, could otherwise amplify latency and, at this layer, cost. Need: bound one serverless tool execution. Circuit breaker around failing tool solves another problem.
Answer C is incorrect because Explicit stopping condition — in this scenario, terminates iterative reasoning or tool-use, in this case, loops when a defined completion,, in this case, risk, or iteration condition is, in this case, reached. Useful when an autonomous loop must not, at this layer, continue indefinitely. Need: bound one serverless tool execution. Explicit stopping condition solves another problem.
Answer D is incorrect because Step Functions retry with backoff — in this scenario, retries transient workflow steps according, in this case, to explicit retry policy and, in this case, delay rather than blindly repeating, in this case, immediately. Useful when a recoverable dependency error should, at this layer, be retried in a controlled, at this layer, way. Need: bound one serverless tool execution. Step Functions retry with backoff solves another problem.
Question 6
An agent may summarize an S3 object but must not have permission to delete or modify the bucket. Which approach should the developer use? Choose ONE.
- IAM least-privilege boundary
- MCP client library
- Tool allowlist
- Security groups for endpoint access
Correct Answer: A
Correct Answer
Answer A is correct because IAM least-privilege boundary — in this scenario, limits the agent or tool, in this case, role to only the AWS, in this case, actions and resources required for, in this case, the task. Useful when resource access must be constrained, at this layer, independently of model behavior. Need: enforce AWS resource permissions independently of the model. IAM least-privilege boundary matches it directly.
Incorrect Answers
Answer B is incorrect because MCP client library — for this design, gives agents a consistent client-side, for this workload, mechanism for discovering and invoking, for this workload, MCP-exposed tools. Fits when multiple MCP servers should be, for the stated need, accessed through a common client, for the stated need, pattern. Constraint: enforce AWS resource permissions independently of the model. MCP client library targets another layer.
Answer C is incorrect because Tool allowlist — in this scenario, restricts the agent to an, in this case, approved set of callable operations, in this case, instead of exposing every available, in this case, action. Useful when the risk boundary is which, at this layer, tools the model may invoke. Need: enforce AWS resource permissions independently of the model. Tool allowlist solves another problem.
Answer D is incorrect because Security groups for endpoint access — in this scenario, restrict network flows to the, in this case, intended sources and destinations around, in this case, VPC resources or endpoints. Useful when private connectivity exists but network-level, at this layer, reachability still needs narrowing. Need: enforce AWS resource permissions independently of the model. Security groups for endpoint access solves another problem.
Question 7
A pricing API has failed repeatedly for five minutes; the agent should stop hammering it and fall back to cached estimates until health recovers. What should the team implement? Choose ONE.
- Step Functions retry with backoff
- Circuit breaker around failing tool
- Idempotent tool operation
- Lambda timeout guard
Correct Answer: B
Correct Answer
Answer B is correct because A circuit breaker stops calls after repeated failures and keeps them suppressed during a cooldown, allowing the application to use cached estimates while the pricing API recovers. Retries alone would continue spending capacity against an unhealthy dependency. This directly matches the requirement to stop hammering a persistently failing service.
Incorrect Answers
Answer A is incorrect because Step Functions retry with backoff spaces out attempts after transient failures and can cap the number of retries. Here, the pricing API has been failing continuously and the application should stop calling it for a cooldown period. A circuit breaker, not continued retry scheduling, provides that suppression behavior.
Answer C is incorrect because Idempotent tool operation — in this scenario, uses an idempotency token or, in this case, state check so a retried, in this case, action does not create duplicate, in this case, side effects. Useful when workflow retries could otherwise repeat, at this layer, a charge, ticket, or update. Need: prevent repeated calls to a persistently failing dependency. Idempotent tool operation solves another problem.
Answer D is incorrect because Lambda timeout guard — at this decision point, enforces bounded execution around a, within this design, tool or model-supporting function so, within this design, stalled work cannot run indefinitely. Typical use: a lightweight serverless action needs, in the described flow, a clear execution timeout. Goal: prevent repeated calls to a persistently failing dependency. Lambda timeout guard does not provide it.
Question 8
Simple classification requests should use a fast economical model, while complex synthesis requests should use a more capable model. Which option most directly addresses the requirement? Choose ONE.
- Model-selection framework
- Specialist-model routing
- AWS Agent Squad multi-agent coordination
- Custom result aggregation
Correct Answer: B
Correct Answer
Answer B is correct because Specialist-model routing — at this decision point, routes a request to a, within this design, model selected for the task’s, within this design, capability, latency, or quality needs. Typical use: different task classes benefit from, in the described flow, different model strengths. Goal: route task classes to different model capabilities. Specialist-model routing provides that behavior.
Incorrect Answers
Answer A is incorrect because Model-selection framework — in this scenario, applies policy or scoring logic, in this case, to choose among candidate models, in this case, at runtime. Useful when the application must choose a, at this layer, model based on measurable request, at this layer, characteristics. Need: route task classes to different model capabilities. Model-selection framework solves another problem.
Answer C is incorrect because AWS Agent Squad multi-agent coordination — for this design, coordinates multiple specialized agents so, for this workload, each can contribute a bounded, for this workload, capability to a larger task. Fits when a workflow benefits from several, for the stated need, specialist agents rather than one, for the stated need, monolithic agent. Constraint: route task classes to different model capabilities. AWS Agent Squad multi-agent coordination targets another layer.
Answer D is incorrect because Custom result aggregation — for this design, combines outputs from multiple model, for this workload, or agent workers using explicit, for this workload, application logic. Fits when several parallel specialist results must, for the stated need, be reconciled into one answer. Constraint: route task classes to different model capabilities. Custom result aggregation targets another layer.
Question 9
Three specialist agents produce risk, cost, and compliance findings that must be merged using deterministic precedence rules. Which design best satisfies this requirement? Choose ONE.
- Custom result aggregation
- Specialist-model routing
- MCP client library
- AWS Agent Squad multi-agent coordination
Correct Answer: A
Correct Answer
Answer A is correct because Custom result aggregation — at this decision point, combines outputs from multiple model, within this design, or agent workers using explicit, within this design, application logic. Typical use: several parallel specialist results must, in the described flow, be reconciled into one answer. Goal: combine several specialist results using application-defined logic. Custom result aggregation provides that behavior.
Incorrect Answers
Answer B is incorrect because Specialist-model routing — under these constraints, routes a request to a, at this layer, model selected for the task’s, at this layer, capability, latency, or quality needs. Choose it when different task classes benefit from, under these constraints, different model strengths. Requirement: combine several specialist results using application-defined logic. Specialist-model routing is not the direct fit.
Answer C is incorrect because MCP client library — at this decision point, gives agents a consistent client-side, within this design, mechanism for discovering and invoking, within this design, MCP-exposed tools. Typical use: multiple MCP servers should be, in the described flow, accessed through a common client, in the described flow, pattern. Goal: combine several specialist results using application-defined logic. MCP client library does not provide it.
Answer D is incorrect because AWS Agent Squad multi-agent coordination — at this decision point, coordinates multiple specialized agents so, within this design, each can contribute a bounded, within this design, capability to a larger task. Typical use: a workflow benefits from several, in the described flow, specialist agents rather than one, in the described flow, monolithic agent. Goal: combine several specialist results using application-defined logic. AWS Agent Squad multi-agent coordination does not provide it.
Question 10
A platform must choose an FM dynamically from latency, quality, and request-type scores rather than a hard-coded if/else list. What is the most appropriate implementation choice? Choose ONE.
- Specialist-model routing
- Custom result aggregation
- Standardized function schema
- Model-selection framework
Correct Answer: D
Correct Answer
Answer D is correct because Model-selection framework — for this design, applies policy or scoring logic, for this workload, to choose among candidate models, for this workload, at runtime. Fits when the application must choose a, for the stated need, model based on measurable request, for the stated need, characteristics. Constraint: select a model by explicit runtime policy. Model-selection framework is the closest fit.
Incorrect Answers
Answer A is incorrect because Specialist-model routing — for the described workload, routes a request to a, for the stated need, model selected for the task’s, for the stated need, capability, latency, or quality needs. Best suited when different task classes benefit from, for this decision, different model strengths. Deciding factor: select a model by explicit runtime policy. Specialist-model routing addresses something else.
Answer B is incorrect because Custom result aggregation — under these constraints, combines outputs from multiple model, at this layer, or agent workers using explicit, at this layer, application logic. Choose it when several parallel specialist results must, under these constraints, be reconciled into one answer. Requirement: select a model by explicit runtime policy. Custom result aggregation is not the direct fit.
Answer C is incorrect because Standardized function schema — for this design, defines tool names, parameters, types,, for this workload, and required fields consistently so, for this workload, the model can call tools, for this workload, reliably. Fits when tool invocation errors are caused, for the stated need, by ambiguous or inconsistent function, for the stated need, definitions. Constraint: select a model by explicit runtime policy. Standardized function schema targets another layer.
Question 11
A loan-assistance agent may draft a recommendation but cannot submit the final action until an authorized employee approves it. Which approach should the developer use? Choose ONE.
- Lambda parameter validation
- Human approval in Step Functions
- API Gateway feedback endpoint
- RETURN_CONTROL action group
Correct Answer: B
Correct Answer
Answer B is correct because Human approval in Step Functions — in this scenario, pauses an orchestrated workflow for, in this case, an explicit human review or, in this case, approval before continuing. Useful when high-impact actions require human judgment, at this layer, before execution. Need: insert a mandatory human approval gate before a high-impact action. Human approval in Step Functions matches it directly.
Incorrect Answers
Answer A is incorrect because Lambda parameter validation — in this scenario, validates and normalizes tool inputs, in this case, before business logic or downstream, in this case, API execution. Useful when tool calls need defensive checks, at this layer, before side effects occur. Need: insert a mandatory human approval gate before a high-impact action. Lambda parameter validation solves another problem.
Answer C is incorrect because API Gateway feedback endpoint — for this design, provides an HTTP interface through, for this workload, which applications can submit structured, for this workload, human feedback or approval data. Fits when the workflow needs an external, for the stated need, feedback collection endpoint. Constraint: insert a mandatory human approval gate before a high-impact action. API Gateway feedback endpoint targets another layer.
Answer D is incorrect because RETURN_CONTROL hands the agent’s selected action and parameters back to the calling application for external execution. That can let an application implement its own controls, but it does not itself impose a mandatory approval checkpoint. The requirement is to pause the workflow until an authorized employee approves the action.
Question 12
A web console needs a secure HTTP endpoint where reviewers can submit approve/reject decisions and comments back to the workflow. What should the team implement? Choose ONE.
- API Gateway feedback endpoint
- Lambda timeout guard
- Human approval in Step Functions
- MCP client library
Correct Answer: A
Correct Answer
Answer A is correct because API Gateway feedback endpoint — at this decision point, provides an HTTP interface through, within this design, which applications can submit structured, within this design, human feedback or approval data. Typical use: the workflow needs an external, in the described flow, feedback collection endpoint. Goal: collect structured reviewer feedback from an external UI. API Gateway feedback endpoint provides that behavior.
Incorrect Answers
Answer B is incorrect because Lambda timeout guard — under these constraints, enforces bounded execution around a, at this layer, tool or model-supporting function so, at this layer, stalled work cannot run indefinitely. Choose it when a lightweight serverless action needs, under these constraints, a clear execution timeout. Requirement: collect structured reviewer feedback from an external UI. Lambda timeout guard is not the direct fit.
Answer C is incorrect because Human approval in Step Functions — for this design, pauses an orchestrated workflow for, for this workload, an explicit human review or, for this workload, approval before continuing. Fits when high-impact actions require human judgment, for the stated need, before execution. Constraint: collect structured reviewer feedback from an external UI. Human approval in Step Functions targets another layer.
Answer D is incorrect because MCP client library — under these constraints, gives agents a consistent client-side, at this layer, mechanism for discovering and invoking, at this layer, MCP-exposed tools. Choose it when multiple MCP servers should be, under these constraints, accessed through a common client, under these constraints, pattern. Requirement: collect structured reviewer feedback from an external UI. MCP client library is not the direct fit.
Question 13
An agent tool performs a short inventory lookup and simple transformation with no persistent process requirements. Which option most directly addresses the requirement? Choose ONE.
- AWS Agent Squad multi-agent coordination
- Lambda action implementation
- Application-managed durable state
- MCP server on Amazon ECS
Correct Answer: B
Correct Answer
Answer B is correct because Lambda action implementation — at this decision point, runs bounded custom business logic, within this design, for an agent tool or, within this design, action group in a serverless, within this design, function. Typical use: a tool action is lightweight,, in the described flow, stateless, and fits Lambda execution, in the described flow, constraints. Goal: run lightweight stateless business logic with minimal operations. Lambda action implementation provides that behavior.
Incorrect Answers
Answer A is incorrect because AWS Agent Squad multi-agent coordination — under these constraints, coordinates multiple specialized agents so, at this layer, each can contribute a bounded, at this layer, capability to a larger task. Choose it when a workflow benefits from several, under these constraints, specialist agents rather than one, under these constraints, monolithic agent. Requirement: run lightweight stateless business logic with minimal operations. AWS Agent Squad multi-agent coordination is not the direct fit.
Answer C is incorrect because Application-managed durable state — in this scenario, stores durable workflow/session state outside, in this case, the model so progress survives, in this case, process restarts. Useful when a long-running agent task must, at this layer, resume reliably after infrastructure interruption. Need: run lightweight stateless business logic with minimal operations. Application-managed durable state solves another problem.
Answer D is incorrect because MCP server on Amazon ECS — in this scenario, hosts an MCP server in, in this case, containers suitable for longer-lived, more, in this case, complex, or dependency-heavy tool services. Useful when the tool server needs persistent, at this layer, process behavior or containerized dependencies. Need: run lightweight stateless business logic with minimal operations. MCP server on Amazon ECS solves another problem.
Question 14
The application must receive the action name and parameters from an agent and invoke an on-premises system itself because AWS cannot call that system directly. Which design best satisfies this requirement? Choose ONE.
- MCP client library
- Lambda action implementation
- RETURN_CONTROL action group
- API Gateway feedback endpoint
Correct Answer: C
Correct Answer
Answer C is correct because RETURN_CONTROL returns the selected action name and parameters to the calling application instead of invoking a Lambda function for fulfillment. Because only the application can reach the on-premises system, the caller can execute the operation through its existing network path and then return the result to the agent.
Incorrect Answers
Answer A is incorrect because MCP client library — for the described workload, gives agents a consistent client-side, for the stated need, mechanism for discovering and invoking, for the stated need, MCP-exposed tools. Best suited when multiple MCP servers should be, for this decision, accessed through a common client, for this decision, pattern. Deciding factor: return the intended action to the caller for application-managed execution. MCP client library addresses something else.
Answer B is incorrect because Lambda action implementation — under these constraints, runs bounded custom business logic, at this layer, for an agent tool or, at this layer, action group in a serverless, at this layer, function. Choose it when a tool action is lightweight,, under these constraints, stateless, and fits Lambda execution, under these constraints, constraints. Requirement: return the intended action to the caller for application-managed execution. Lambda action implementation is not the direct fit.
Answer D is incorrect because API Gateway feedback endpoint — under these constraints, provides an HTTP interface through, at this layer, which applications can submit structured, at this layer, human feedback or approval data. Choose it when the workflow needs an external, under these constraints, feedback collection endpoint. Requirement: return the intended action to the caller for application-managed execution. API Gateway feedback endpoint is not the direct fit.
Question 15
Several tools fail because equivalent parameters are named differently and required fields are unclear to the model. What is the most appropriate implementation choice? Choose ONE.
- Standardized function schema
- Lambda parameter validation
- Model-selection framework
- Tool allowlist
Correct Answer: A
Correct Answer
Answer A is correct because Standardized function schema — at this decision point, defines tool names, parameters, types,, within this design, and required fields consistently so, within this design, the model can call tools, within this design, reliably. Typical use: tool invocation errors are caused, in the described flow, by ambiguous or inconsistent function, in the described flow, definitions. Goal: make tool contracts unambiguous before invocation. Standardized function schema provides that behavior.
Incorrect Answers
Answer B is incorrect because Lambda parameter validation — for this design, validates and normalizes tool inputs, for this workload, before business logic or downstream, for this workload, API execution. Fits when tool calls need defensive checks, for the stated need, before side effects occur. Constraint: make tool contracts unambiguous before invocation. Lambda parameter validation targets another layer.
Answer C is incorrect because Model-selection framework — at this decision point, applies policy or scoring logic, within this design, to choose among candidate models, within this design, at runtime. Typical use: the application must choose a, in the described flow, model based on measurable request, in the described flow, characteristics. Goal: make tool contracts unambiguous before invocation. Model-selection framework does not provide it.
Answer D is incorrect because Tool allowlist — for this design, restricts the agent to an, for this workload, approved set of callable operations, for this workload, instead of exposing every available, for this workload, action. Fits when the risk boundary is which, for the stated need, tools the model may invoke. Constraint: make tool contracts unambiguous before invocation. Tool allowlist targets another layer.
Question 16
A tool receives a quantity and account identifier from the model. Invalid quantities and malformed IDs must be rejected before any transaction is attempted. Which approach should the developer use? Choose ONE.
- Standardized function schema
- Idempotent tool operation
- Lambda parameter validation
- IAM least-privilege boundary
Correct Answer: C
Correct Answer
Answer C is correct because Lambda parameter validation — at this decision point, validates and normalizes tool inputs, within this design, before business logic or downstream, within this design, API execution. Typical use: tool calls need defensive checks, in the described flow, before side effects occur. Goal: validate tool inputs before side effects. Lambda parameter validation provides that behavior.
Incorrect Answers
Answer A is incorrect because Standardized function schema — under these constraints, defines tool names, parameters, types,, at this layer, and required fields consistently so, at this layer, the model can call tools, at this layer, reliably. Choose it when tool invocation errors are caused, under these constraints, by ambiguous or inconsistent function, under these constraints, definitions. Requirement: validate tool inputs before side effects. Standardized function schema is not the direct fit.
Answer B is incorrect because Idempotent tool operation — for this design, uses an idempotency token or, for this workload, state check so a retried, for this workload, action does not create duplicate, for this workload, side effects. Fits when workflow retries could otherwise repeat, for the stated need, a charge, ticket, or update. Constraint: validate tool inputs before side effects. Idempotent tool operation targets another layer.
Answer D is incorrect because IAM least-privilege boundary — for this design, limits the agent or tool, for this workload, role to only the AWS, for this workload, actions and resources required for, for this workload, the task. Fits when resource access must be constrained, for the stated need, independently of model behavior. Constraint: validate tool inputs before side effects. IAM least-privilege boundary targets another layer.
Question 17
A team needs a small MCP tool service that is stateless, invoked sporadically, and has no native library dependencies. What should the team implement? Choose ONE.
- Lambda action implementation
- MCP client library
- Stateless MCP server on Lambda
- MCP server on Amazon ECS
Correct Answer: C
Correct Answer
Answer C is correct because Stateless MCP server on Lambda — in this scenario, hosts lightweight, stateless MCP tool, in this case, access in Lambda with minimal, in this case, server management. Useful when the MCP workload is short-lived,, at this layer, stateless, and event-driven. Need: host a lightweight stateless MCP server. Stateless MCP server on Lambda matches it directly.
Incorrect Answers
Answer A is incorrect because Lambda action implementation — for the described workload, runs bounded custom business logic, for the stated need, for an agent tool or, for the stated need, action group in a serverless, for the stated need, function. Best suited when a tool action is lightweight,, for this decision, stateless, and fits Lambda execution, for this decision, constraints. Deciding factor: host a lightweight stateless MCP server. Lambda action implementation addresses something else.
Answer B is incorrect because MCP client library — in the current architecture, gives agents a consistent client-side, in the described flow, mechanism for discovering and invoking, in the described flow, MCP-exposed tools. Applies where multiple MCP servers should be, in this case, accessed through a common client, in this case, pattern. Target outcome: host a lightweight stateless MCP server. MCP client library misses that outcome.
Answer D is incorrect because MCP server on Amazon ECS — for this design, hosts an MCP server in, for this workload, containers suitable for longer-lived, more, for this workload, complex, or dependency-heavy tool services. Fits when the tool server needs persistent, for the stated need, process behavior or containerized dependencies. Constraint: host a lightweight stateless MCP server. MCP server on Amazon ECS targets another layer.
Question 18
An MCP tool service needs long-lived container processes, native packages, and predictable CPU/memory for complex analysis. Which option most directly addresses the requirement? Choose ONE.
- AWS Agent Squad multi-agent coordination
- Stateless MCP server on Lambda
- Step Functions orchestration
- MCP server on Amazon ECS
Correct Answer: D
Correct Answer
Answer D is correct because MCP server on Amazon ECS — at this decision point, hosts an MCP server in, within this design, containers suitable for longer-lived, more, within this design, complex, or dependency-heavy tool services. Typical use: the tool server needs persistent, in the described flow, process behavior or containerized dependencies. Goal: host a dependency-heavy longer-lived MCP server. MCP server on Amazon ECS provides that behavior.
Incorrect Answers
Answer A is incorrect because AWS Agent Squad multi-agent coordination — for the described workload, coordinates multiple specialized agents so, for the stated need, each can contribute a bounded, for the stated need, capability to a larger task. Best suited when a workflow benefits from several, for this decision, specialist agents rather than one, for this decision, monolithic agent. Deciding factor: host a dependency-heavy longer-lived MCP server. AWS Agent Squad multi-agent coordination addresses something else.
Answer B is incorrect because Stateless MCP server on Lambda — for this design, hosts lightweight, stateless MCP tool, for this workload, access in Lambda with minimal, for this workload, server management. Fits when the MCP workload is short-lived,, for the stated need, stateless, and event-driven. Constraint: host a dependency-heavy longer-lived MCP server. Stateless MCP server on Lambda targets another layer.
Answer C is incorrect because Step Functions orchestration — for this design, models multi-step workflow control, branching,, for this workload, retries, approvals, and stopping conditions, for this workload, explicitly. Fits when an AI workflow needs deterministic, for the stated need, orchestration around model or tool, for the stated need, calls. Constraint: host a dependency-heavy longer-lived MCP server. Step Functions orchestration targets another layer.
Question 19
One agent must discover and call tools exposed by three separate MCP servers using the same invocation pattern. Which design best satisfies this requirement? Choose ONE.
- Custom result aggregation
- MCP client library
- Standardized function schema
- MCP for agent-tool interoperability
Correct Answer: B
Correct Answer
Answer B is correct because MCP client library — for this operational need, gives agents a consistent client-side, under these constraints, mechanism for discovering and invoking, under these constraints, MCP-exposed tools. Architects use it when multiple MCP servers should be, for this workload, accessed through a common client, for this workload, pattern. Key need: use a common client mechanism across MCP servers. MCP client library aligns with it.
Incorrect Answers
Answer A is incorrect because Custom result aggregation — for the described workload, combines outputs from multiple model, for the stated need, or agent workers using explicit, for the stated need, application logic. Best suited when several parallel specialist results must, for this decision, be reconciled into one answer. Deciding factor: use a common client mechanism across MCP servers. Custom result aggregation addresses something else.
Answer C is incorrect because A standardized function schema makes the name, parameters, types, and required fields of one tool predictable to the model. It does not give the agent a common discovery and invocation protocol across independent MCP servers. The requirement is interoperability at the client/protocol layer, so an MCP client mechanism is the better fit.
Answer D is incorrect because MCP for agent-tool interoperability — at this decision point, uses the Model Context Protocol, within this design, to expose tools through a, within this design, consistent agent-to-tool interface. Typical use: tools should be reachable through, in the described flow, a standardized protocol instead of, in the described flow, bespoke integrations. Goal: use a common client mechanism across MCP servers. MCP for agent-tool interoperability does not provide it.
Question 20
A ReAct-style workflow keeps asking the same search tool even after sufficient evidence has been collected. What is the most appropriate implementation choice? Choose ONE.
- Tool allowlist
- Circuit breaker around failing tool
- Lambda timeout guard
- Explicit stopping condition
Correct Answer: D
Correct Answer
Answer D is correct because Explicit stopping condition — for this design, terminates iterative reasoning or tool-use, for this workload, loops when a defined completion,, for this workload, risk, or iteration condition is, for this workload, reached. Fits when an autonomous loop must not, for the stated need, continue indefinitely. Constraint: terminate an autonomous loop when completion criteria are met. Explicit stopping condition is the closest fit.
Incorrect Answers
Answer A is incorrect because Tool allowlist — at this decision point, restricts the agent to an, within this design, approved set of callable operations, within this design, instead of exposing every available, within this design, action. Typical use: the risk boundary is which, in the described flow, tools the model may invoke. Goal: terminate an autonomous loop when completion criteria are met. Tool allowlist does not provide it.
Answer B is incorrect because A circuit breaker is designed for an unhealthy dependency: after repeated failures it suppresses additional calls until recovery conditions are met. The search tool here is not failing; the agent simply continues after it already has enough evidence. An explicit stopping condition should terminate the reasoning-and-tool loop when completion criteria are satisfied.
Answer C is incorrect because Lambda timeout guard — for the described workload, enforces bounded execution around a, for the stated need, tool or model-supporting function so, for the stated need, stalled work cannot run indefinitely. Best suited when a lightweight serverless action needs, for this decision, a clear execution timeout. Deciding factor: terminate an autonomous loop when completion criteria are met. Lambda timeout guard addresses something else.
