sast-eslint-security
Fail
Audited by Gen Agent Trust Hub on Feb 14, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (HIGH): The skill ingests untrusted code from target-path. Malicious instructions within comments in the scanned code can be reflected in the scan results and mislead the agent. Evidence: 1. Ingestion: target-path content. 2. Boundaries: None. 3. Capabilities: Shell execution, file writing. 4. Sanitization: None.
- Command Execution (MEDIUM): The skill executes npx eslint via shell. If the target-path is not sanitized, an attacker could execute arbitrary commands via shell injection.
- External Downloads (LOW): Uses npx and npm to install/run linting tools. While the sources are trusted, runtime downloads via npx lack version pinning by default.
Recommendations
- AI detected serious security threats
Audit Metadata