skills/lijigang/ljg-skills/ljg-plain/Gen Agent Trust Hub

ljg-plain

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the system 'date' command to generate identifiers and timestamps for file management. This is a benign use of system utilities for organization purposes.
  • [EXTERNAL_DOWNLOADS]: The skill includes functionality to fetch content from external URLs and search engines to gather source material for simplification.
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration attempts were detected. File system access is limited to a specific user directory (~/Documents/notes/) for saving generated notes.
  • [PROMPT_INJECTION]: The skill processes untrusted data from external sources (Ingestion points: WebFetch, WebSearch). No explicit boundary markers are used, but the 'Red Lines' constraints enforce a transformation process that prevents direct execution of embedded instructions. Capabilities are restricted to metadata generation and local note storage (Capability inventory: 'date' command, file-write).
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 07:22 AM