security-assessment

Warn

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill is designed to perform direct shell execution of bash scripts (e.g., detect-tools.sh, run-tool.sh, validate-output.sh) to coordinate assessment tasks and analyze tool logs.
  • [REMOTE_CODE_EXECUTION]: The skill implements an auto-detection protocol for its root directory (SUPERHACKERS_ROOT) that searches multiple paths, including the current working directory via $(pwd)/superhackers. This creates a vulnerability where a malicious directory structure in a project could hijack the search path and cause the agent to execute untrusted local scripts.
  • [EXTERNAL_DOWNLOADS]: The documentation specifies a dependency on the cvss Python library and several external security binaries (such as nmap, nuclei, and sqlmap) which must be present on the host system to function correctly.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 20, 2026, 10:17 AM
Security Audit — agent-trust-hub — security-assessment