dead-code-detector
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [Command Execution] (MEDIUM): The skill instructs the agent to execute a local JavaScript file 'find-dead-code.js'. The content of this script is not provided, representing an unverified execution path on the host system.
- [Prompt Injection] (HIGH): The skill has a high-risk surface for indirect prompt injection as it processes untrusted codebase content to drive destructive cleanup actions. Mandatory Evidence Chain: 1. Ingestion points: Project root directory and all source files via the scanner script. 2. Boundary markers: Absent; the scanner output and manual investigation commands do not use delimiters or instructions to ignore embedded commands. 3. Capability inventory: Execution of local scripts, ripgrep searches, and the capability to delete files or uninstall packages via the cleanup workflow. 4. Sanitization: Absent. A malicious actor could craft code patterns that appear to be 'dead code' to the scanner, tricking the agent into deleting critical security checks or business logic.
Recommendations
- AI detected serious security threats
Audit Metadata