stock-report

Warn

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The SKILL.md file passes the $ARGUMENTS variable directly to a bash command (python3 .../generate_report.py $ARGUMENTS) without quoting or validation, which could allow shell injection if the input contains metacharacters. Additionally, generate_report.py performs dynamic path manipulation by using sys.path.insert to add parent directories to the module search path at runtime.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection from external financial data. 1. Ingestion points: Ticker symbols from user input and stock detail data fetched via the Yahoo Finance client. 2. Boundary markers: Absent; the script prints the retrieved data directly to the agent's context as plain Markdown. 3. Capability inventory: The skill utilizes Bash(python3) which provides broad command execution capabilities. 4. Sanitization: There is no evidence of sanitization or validation of the input symbol or the content returned from the external stock data provider.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 3, 2026, 06:43 PM