announcement-analysis
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to interact with the
finxCLI tool to query thegildata-aidatadatabase for financial indicators, valuation metrics, and institutional ratings. This is a standard and expected functional use of platform-integrated financial tools. - [PROMPT_INJECTION]: The skill processes untrusted external data from user-uploaded files and web search results, presenting a surface for indirect prompt injection. This is a functional requirement for the skill's primary purpose of document analysis, and the instructions provide a rigid reporting framework that acts as a natural constraint on agent behavior.
Audit Metadata