company-current-gtm-analysis
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's primary function is to ingest and synthesize untrusted data from various external sources such as blogs, Reddit, and LinkedIn posts.
- Ingestion points: Untrusted data enters the agent context through
WebFetch,WebSearch, and the output of the LinkedIn scraper script. - Boundary markers: The instructions do not provide delimiters or "ignore embedded instructions" warnings for the agent when processing this external content.
- Capability inventory: The skill has the ability to write to the local filesystem (creating research reports in
clients/) and perform further network operations. - Sanitization: There is no mention of sanitizing or escaping external content before it is interpolated into the synthesis and scoring prompts.
- [EXTERNAL_DOWNLOADS]: The skill makes extensive network requests to the public internet to download content from blogs, social media platforms, and review sites using
WebFetchandWebSearchtools. - [COMMAND_EXECUTION]: The skill uses a shell command to execute a local Python script (
scrape_linkedin_posts.py) for scraping LinkedIn data, which involves passing external URLs as arguments to the script.
Audit Metadata