mcp-builder

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill includes MCP CallTool handlers that fetch and return data from external public APIs (e.g., "Example 2: API Integration" uses fetch(https://api.weather.com/v1/${city})) and recommends building tools for services like GitHub/Slack (github_list_repos, slack_list_channels) which would return user-generated, untrusted third-party content for the agent to read and interpret.