project-scanner
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (LOW): The skill instructs the agent to execute a local script 'scan.ts' using 'npx tsx'. This executes code from the local filesystem that is not provided in the skill definition. Additionally, the documentation contains a hardcoded absolute file path 'C:\Users\BLogic\...' which exposes a local username.
- [PROMPT_INJECTION] (LOW): Indirect Prompt Injection Surface: (1) Ingestion points: The tool scans project source code via the '--path' argument. (2) Boundary markers: None specified in the skill instructions to delimit code from instructions. (3) Capability inventory: Executes shell commands (npx tsx) and reads local JSON output files. (4) Sanitization: No sanitization or escaping of the scanned source code is mentioned before processing.
Audit Metadata