scope
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell commands (cat, mkdir, jq, mv) to create and update the .pentest/scope.json file.
- [PROMPT_INJECTION]: The skill processes user input for scope management, creating an indirect injection surface. 1. Ingestion points: Target strings provided via $ARGUMENTS in SKILL.md. 2. Boundary markers: Manual user confirmation is required for all write operations. 3. Capability inventory: Local file system access and shell command execution. 4. Sanitization: Employs jq --arg which safely escapes variables for JSON manipulation, preventing injection into the jq processor.
Audit Metadata