The Agent Skills Directory

Indirect Prompt Injection (MEDIUM): The skill is designed to ingest and process data from external Google Ads and Analytics accounts, which constitutes an untrusted data surface.
Ingestion points: Data enters the agent's context through tools like GOOGLE_ANALYTICS_RUN_REPORT and GOOGLE_ANALYTICS_LIST_PROPERTIES (campaign names, account descriptions, etc.).
Boundary markers: Absent. The skill does not provide delimiters or instructions for the agent to disregard natural language instructions that might be embedded in campaign names or metadata.
Capability inventory: The provided tools are primarily read-only (reporting and listing), which limits the immediate risk to information disclosure or reasoning bias rather than direct system compromise.
Sanitization: There is no evidence of sanitization or filtering of the data retrieved from the API.
External Downloads (LOW): The skill requires the configuration of a remote MCP server at https://rube.app/mcp.
Evidence: The setup instructions explicitly require adding this third-party URL as a server endpoint.
Trust Status: Neither rube.app nor composio.dev are included in the provided list of trusted external sources. Users should verify the security and privacy practices of the provider before connecting their Google accounts.

googleads-automation