geo-audit
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external websites and third-party platforms, creating a potential indirect prompt injection surface. 1. Ingestion points: Target URLs (homepage, robots.txt, sitemaps) and third-party services (Wikipedia, Wikidata, LinkedIn, Reddit, YouTube, Crunchbase). 2. Boundary markers: Explicit tags and instructions to treat external content as data. 3. Capability inventory: Network fetches, subagent spawning, file writing, and document conversion. 4. Sanitization: Logic to detect and log 'Prompt Injection Attempt Detected' strings.
- [EXTERNAL_DOWNLOADS]: The skill fetches metadata and brand signals from reputable well-known platforms including Wikipedia, Wikidata, LinkedIn, Reddit, YouTube, and Crunchbase.
- [COMMAND_EXECUTION]: The skill instructs the agent to use conversion utilities like pandoc to export reports into PDF and Word formats.
Audit Metadata