google-indexing-service
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute the
giscommand-line tool to perform indexing requests. Evidence: Examples inSKILL.mdsuch asgis example.com --path "$GSC_CREDENTIALS_PATH".\n- [DATA_EXFILTRATION]: The skill is designed to work with sensitive Google Cloud service account credentials (JSON format) accessed via theGSC_CREDENTIALS_PATHenvironment variable. While this is a standard configuration for the tool, it involves processing highly sensitive authentication data.\n- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by ingesting untrusted data (domains and URLs) from the user and interpolating it directly into command arguments for thegistool.\n - Ingestion points: User-provided domain and specific page URLs referenced in the core workflow sections of
SKILL.md.\n - Boundary markers: None identified; user-supplied strings are passed directly to the shell.\n
- Capability inventory: Execution of shell commands via the
gisutility.\n - Sanitization: No instructions are provided to sanitize or validate the external input before it is used in the command execution.
Audit Metadata