company-current-gtm-analysis
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a Python script located at
skills/linkedin-profile-post-scraper/scripts/scrape_linkedin_posts.pyto retrieve social media data. This is a core functional requirement of the research process. - [EXTERNAL_DOWNLOADS]: Extensive web fetching and searching are performed across domains like company blogs, career pages, and review sites (G2, Capterra). These operations are used solely to aggregate public information for the report.
- [PROMPT_INJECTION]: The skill ingests untrusted content from external web sources in Phase 2. There are no explicit boundary markers or sanitization steps mentioned. However, the risk is mitigated by the highly structured template used in Phase 3 and Phase 4, which directs the agent to organize data rather than execute instructions found within it.
Audit Metadata