Mock MCP for Port Demos

This repository provides mock observability context for Port's Agentic Engineering Platform demos.

Usage with GitMCP

Access via GitMCP URLs:

• Datadog: port-labs.gitmcp.io/mock-mcp/datadog
• GitHub: port-labs.gitmcp.io/mock-mcp/github
• NewRelic: port-labs.gitmcp.io/mock-mcp/newrelic
• AWS: port-labs.gitmcp.io/mock-mcp/aws

How It Works

Each subdirectory contains an llms.txt that instructs AI agents how to behave when simulating that vendor's tools.

Port sends context about entities (services, incidents) and hints about what kind of response to generate. The AI reads the relevant llms.txt and generates appropriate mock responses.

Hint System

Port can include hints in the context:

• found_issue - Generate data showing a problem
• nothing_found - Generate healthy/clean data
• not_my_domain - Explain this tool can't help