quality-gates
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- PROMPT_INJECTION (MEDIUM): The skill directs the agent to read and update 'requirements.md'. Because this file has been flagged by automated security scanners for containing a malicious URL, following these instructions exposes the agent to potential indirect prompt injection or malicious data processing.
- COMMAND_EXECUTION (LOW): The skill requires the agent to execute various local shell commands such as 'pnpm', 'eslint', and 'tsc'. These represent a baseline capability for command execution that could be leveraged if the agent's logic is subverted through the flagged external files.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata