The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the google-business-profile-pp-cli binary and an MCP server. It provides instructions to install these via go install from github.com/mvanhorn/printing-press-library and via npx from the @mvanhorn/printing-press-library package on npm. While these are identified as vendor-owned resources, they involve executing code from an external repository.
[DATA_EXFILTRATION]: The CLI tool documents a --deliver flag that supports a webhook:<url> sink. This feature POSTs the command output body directly to a specified URL. Because command results can include sensitive API data, including account IDs, administrator lists, and location details, this provides a direct mechanism for data exfiltration if the agent is manipulated into using a malicious endpoint.
[COMMAND_EXECUTION]: The skill requires the execution of a custom CLI tool with broad permissions over the Google Business Profile API. It includes commands for administrative actions such as create-admins, accounts delete, and my-business-notifications-accounts update-notification-setting.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes and displays data from external API sources.
Ingestion points: Data retrieved from Google Business Profile API resources (accounts, locations, reviews, etc.) via the archive, search, and tail commands.
Boundary markers: No specific delimiters or warnings to ignore embedded instructions were identified in the skill content.
Capability inventory: The agent has access to tools for modifying account data and sending data to external webhooks via the CLI.
Sanitization: No sanitization or validation of the API-derived content is described before it is integrated into the agent's context.

pp-google-business-profile