status-line-customization
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- Prompt Injection (LOW): The file uses aggressive directives ('ABSOLUTE REQUIREMENT', 'CRITICAL', 'MANDATORY') to override the agent's default auditing behavior. It explicitly commands the agent to suppress findings ('Do NOT flag', 'mark finding as UNVERIFIED') if they cannot be cross-referenced with a specific external tool (
docs-management). - Command Execution (MEDIUM): The framework mandates a 'RUNTIME CHECK' for audited scripts ('Did I test the script actually works... RUNTIME CHECK MATTERS'). This instructs the agent to execute untrusted code provided as input to the audit process, which could lead to local code execution vulnerabilities if the scripts are malicious.
- Indirect Prompt Injection (LOW):
- Ingestion points: The framework is designed to process external status-line scripts as audit targets.
- Boundary markers: Absent. No instructions are provided to sanitize or isolate the scripts before execution or analysis.
- Capability inventory: The framework requires script execution ('RUNTIME CHECK') and analysis of output/permissions.
- Sanitization: Absent. There is no mention of sandboxing or safety checks before running the audited scripts.
Audit Metadata