skills/gfargo/skills/strut/Gen Agent Trust Hub

strut

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands locally and on remote VPS hosts via SSH for container management, system auditing, and diagnostic purposes.
  • [EXTERNAL_DOWNLOADS]: The migration wizard includes an option to fetch and run the official Docker installation script from 'https://get.docker.com', which is a well-known and trusted source for infrastructure setup.
  • [DATA_EXFILTRATION]: The skill facilitates the synchronization of environment secrets and SSH keys between local storage and remote VPS hosts as part of its core secret management functionality.
  • [PROMPT_INJECTION]: The tool's ingestion of external configuration files and remote audit data creates a surface for indirect prompt injection. 1. Ingestion points: Reads stack configurations and environment templates from the file system and audit data from remote VPS hosts. 2. Boundary markers: No specific delimiters or instructions are used to isolate potentially untrusted configuration content. 3. Capability inventory: Includes full file system access, network communication via SSH, and arbitrary shell command execution on managed hosts. 4. Sanitization: Uses a dedicated safe parser for environment files to prevent unintended shell expansion of values.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 08:35 AM
Security Audit — agent-trust-hub — strut