kol-discovery
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: Executes a Python script to automate the aggregation, scoring, and ranking of Key Opinion Leaders (KOLs) based on engagement metrics.
- [EXTERNAL_DOWNLOADS]: Fetches data from the LinkedIn post search service via the vendor's official API proxy at app.gooseworks.ai.
- [CREDENTIALS_UNSAFE]: Accesses the GOOSEWORKS_API_KEY and APIFY_API_TOKEN from the execution environment to facilitate authorized requests to the search service.
- [PROMPT_INJECTION]: Processes external content from LinkedIn posts as part of its discovery workflow; while this introduces a theoretical surface for indirect prompt injection, the risk is mitigated by content truncation to 100-200 characters.
Audit Metadata