google-ads-strategy
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to verify if the vendor's CLI tool is installed by executing the
adkit statuscommand within the user's environment. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted data from local project files.
- Ingestion points: Reads and applies content from
ad-process.mdandad-brief.mdfound in the user's project workspace to tailor its recommendations. - Boundary markers: The instructions do not define explicit delimiters or security boundaries to isolate potentially malicious instructions embedded within these configuration files.
- Capability inventory: The ingested data directly influences the generation of ad strategies, keyword mining results, and ad copy.
- Sanitization: There is no mention of validation, filtering, or sanitization being performed on the content of the ingested files.
Audit Metadata