pp-google-search-console
Warn
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install software from external repositories, specifically from npm (@mvanhorn/printing-press) and GitHub (github.com/mvanhorn/printing-press-library).
- [REMOTE_CODE_EXECUTION]: The installation process involves executing remote code directly via 'npx' and 'go install' to provide the necessary command-line environment.
- [DATA_EXFILTRATION]: The CLI tool provides a '--deliver webhook:' flag, which allows the agent to send command outputs, potentially containing private search analytics or site data, to any arbitrary external URL.
- [COMMAND_EXECUTION]: The skill operates by invoking a local binary using the 'Bash' tool, with several commands accepting complex arguments that are passed to the shell.
- [PROMPT_INJECTION]: The 'which' command is designed to resolve natural language capability queries to CLI commands, which creates a surface for indirect prompt injection if untrusted user input is passed directly into the tool without sanitization.
Audit Metadata