investment-checklist
Pass
Audited by Gen Agent Trust Hub on Jul 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using
python3to run a local utility (~/ai-berkshire/tools/financial_rigor.py) for financial verification and scenario analysis. - [EXTERNAL_DOWNLOADS]: The skill uses
WebSearchand theTasktool to fetch live financial data and company news from the internet. - [PROMPT_INJECTION]: The skill exhibits a surface for Indirect Prompt Injection.
- Ingestion points: Untrusted data (financial metrics, CEO names, news) enters the context via
WebSearchand backgroundAgenttasks in Step 2. - Boundary markers: Absent; the skill does not use delimiters or instructions to treat external data as untrusted text.
- Capability inventory: The skill has access to the
Bashtool for command execution and can write files to the home directory (~/巴菲特Checklist-...). - Sanitization: Absent; external data points like
{股价}or{EPS}are directly interpolated into thepython3shell command strings in Step 3.
Audit Metadata