skills/tt-a1i/hermes-agent/spike/Gen Agent Trust Hub

spike

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the terminal to manage directories, check package versions (e.g., pip show), and execute generated prototype scripts.
  • [DYNAMIC_EXECUTION]: The skill writes user-defined logic to files and executes them as independent programs using local runtimes.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external web research and user prompts to generate executable code.
  • Ingestion points: User feasibility questions and external documentation retrieved via web research tools.
  • Boundary markers: None present to delimit or warn against instructions within external content.
  • Capability inventory: Terminal access, file writing, and task delegation.
  • Sanitization: No explicit logic to sanitize or validate the generated code before execution.
  • [EXTERNAL_DOWNLOADS]: The documentation references external projects and suggests installing tools via npx or cloning repositories from GitHub.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 12:48 PM
Security Audit — agent-trust-hub — spike