Integration and APIs

In environment variables or a secret manager, not hard coded in code.

Bounded retries with backoff and jitter, respecting server headers where available.

Request id, endpoint, status code, timing, and a safe summary of inputs and outputs, without leaking secrets.

Because usage spikes and runaway loops can turn into a billing incident quickly.

1. To execute the tool
2. To format the tool for inclusion in agent prompts
3. To validate tool inputs
4. To log tool usage

Correct answer: To format the tool for inclusion in agent prompts

The to_prompt_format method formats the tool's name, parameters, and description so it can be included in the agent's prompt, helping the LLM understand what tools are available.

1. Execute all tasks directly
2. Route requests to specialised agents
3. Store conversation history
4. Handle error logging

Correct answer: Route requests to specialised agents

The supervisor's main role is to route requests to the most appropriate specialised agent and synthesise their responses. It coordinates but does not do the specialised work itself.

1. The agent crashes
2. A helpful error message is returned
3. The tool is automatically created
4. The query is ignored

Correct answer: A helpful error message is returned

Good agent implementations return a helpful error message listing available tools. This allows the agent to correct its mistake in the next iteration.

1. Because it is faster than text
2. Because it provides structured, parseable parameters
3. Because LLMs only output JSON
4. Because it reduces token usage

Correct answer: Because it provides structured, parseable parameters

JSON provides structured, parseable parameters that can be directly passed to Python functions. This is more reliable than trying to parse free-form text.

1. Lower latency for simple tasks
2. Better error handling
3. Agents can hand off directly without central bottleneck
4. Easier to implement

Correct answer: Agents can hand off directly without central bottleneck

In the Swarm pattern, agents hand off directly to each other. This avoids the supervisor as a central bottleneck and allows more natural peer-to-peer collaboration.

1. REST API
2. GraphQL
3. JSON-RPC 2.0
4. gRPC

Correct answer: JSON-RPC 2.0

MCP uses JSON-RPC 2.0 for communication between clients and servers, with transport details defined by the current specification.

1. Better performance
2. Reduced number of total integrations needed
3. Free hosting
4. Automatic testing

Correct answer: Reduced number of total integrations needed

MCP reduces integrations from N apps x M tools to N + M implementations. Each app implements MCP once, each tool implements MCP once.

1. Mobile app development
2. Visual workflow automation with AI
3. Database management
4. Video editing

Correct answer: Visual workflow automation with AI

n8n is a visual workflow automation platform with native AI capabilities. It lets you build AI-powered workflows using a drag-and-drop interface.

1. They load faster
2. They are automatically encrypted
3. They stay out of version control and are easier to rotate
4. They reduce memory usage

Correct answer: They stay out of version control and are easier to rotate

Environment variables keep secrets out of source code and version control. They can be easily changed without code modifications, making key rotation simpler.

1. A type of encryption
2. Increasing wait time between retry attempts
3. A memory management technique
4. A way to compress data

Correct answer: Increasing wait time between retry attempts

Exponential backoff means increasing the wait time between retry attempts (e.g., 1s, 2s, 4s, 8s). This prevents overwhelming servers during outages.