The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data ingested from Google Analytics 4 reports, which represents a potential surface for indirect prompt injection if report dimensions (like page titles, event names, or custom dimensions) contain malicious instructions.
Ingestion points: Data enters the agent context through the outputs of reporting tools such as GOOGLE_ANALYTICS_RUN_REPORT, GOOGLE_ANALYTICS_RUN_PIVOT_REPORT, and GOOGLE_ANALYTICS_RUN_FUNNEL_REPORT (SKILL.md).
Boundary markers: The instructions do not specify the use of delimiters or specific warnings to ignore instructions embedded within the analytics data.
Capability inventory: The skill provides extensive read and list capabilities for Google Analytics accounts and properties, as well as connection management via RUBE_MANAGE_CONNECTIONS (SKILL.md).
Sanitization: There are no explicit instructions for the agent to sanitize or validate the content of the report data before processing or presenting it.
[EXTERNAL_DOWNLOADS]: The skill instructs the user to configure a remote MCP server at https://rube.app/mcp and references documentation at composio.dev. These are legitimate external resources provided by Composio for AI agent toolkits.

google-analytics-automation