mobile-seo
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run local Python scripts to retrieve data from PageSpeed Insights, Chrome User Experience Report (CrUX), and Lighthouse, as well as to publish final reports.
- Evidence: Use of
python scripts/google_client.py,python scripts/dfs_client.py, andpython scripts/report_publish.pyfor metrics collection and report generation. - [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes external HTML content, which is a common ingestion point for instructions embedded in third-party data.
- Ingestion points: External page content is retrieved via the
webfetchtool (SKILL.md). - Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potential instructions embedded in the fetched HTML.
- Capability inventory: The skill has the capability to execute local scripts and generate formatted reports based on the analyzed data.
- Sanitization: No sanitization or filtering of the external HTML content is specified before the comparison analysis occurs.
Audit Metadata