us-stock-analysis
Pass
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: SAFENO_CODE
Full Analysis
- NO_CODE (SAFE): The skill is comprised entirely of Markdown documentation (
SKILL.mdand reference files). It does not include any Python scripts, Node.js packages, or shell commands, significantly reducing the attack surface. - DATA_EXFILTRATION (SAFE): While the skill instructs the agent to fetch market data from external sources (e.g., Yahoo Finance, SEC filings), this is restricted to public financial information. No patterns for accessing sensitive local files (like
~/.aws/credentials) or environment variables were detected. - PROMPT_INJECTION (SAFE): The instructions follow best practices for financial analysis and do not contain patterns intended to bypass agent safety filters or override system prompts.
- COMMAND_EXECUTION (SAFE): There are no instances of subprocess spawning, shell execution, or privilege escalation attempts within the provided files.
Audit Metadata