fuzz
Warn
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDYNAMIC_CONTEXT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_CONTEXT_INJECTION]: The skill uses the
!commandsyntax to run a shell check (test -f) for the existence of local wrapper scripts when the skill is loaded into the agent context. - [COMMAND_EXECUTION]: The skill directly interpolates the user-provided
$ARGUMENTSinto shell commands for tools likegobuster,ffuf, andnikto, as well as custom wrapper scripts. This pattern is vulnerable to command injection if the input is not strictly validated to be a URL. - [INDIRECT_PROMPT_INJECTION]: The skill reads and processes the content of a local project file to verify target scope, creating an indirect injection surface.
- Ingestion points:
SKILL.md(viacat .pentest/scope.json). - Boundary markers: None present to isolate the file content.
- Capability inventory: Shell execution for network scanning and brute-forcing tools.
- Sanitization: No sanitization or validation logic is applied to the content of the scope file.
Audit Metadata