sgc-write-professional-book

Warn

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires the execution of shell commands, specifically uvx sgc-skill-helper and npx lovstudio, to decrypt its instructions and manage licensing.
  • [REMOTE_CODE_EXECUTION]: The skill utilizes runtime package execution tools (uvx and npx) to download and run external code from public registries (PyPI and NPM) to perform its core functions.
  • [EXTERNAL_DOWNLOADS]: The decryption process involves network communication to verify licenses and fetch content, and the skill relies on the availability of external packages fetched at runtime.
  • [PROMPT_INJECTION]: By employing an encrypted loader pattern, the skill prevents the agent and platform security tools from inspecting the actual instructions and logic until the moment of execution, which can be used to bypass static safety filters.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 10, 2026, 12:14 PM
Security Audit — agent-trust-hub — sgc-write-professional-book