skills/steveclarke/dotfiles/humanizer/Gen Agent Trust Hub

humanizer

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional, focusing on linguistic patterns and stylistic improvements without any code or dangerous operations.
  • [DATA_EXFILTRATION]: No network access or data exfiltration patterns were found. The skill includes a link to a trusted Wikipedia page for reference.
  • [REMOTE_CODE_EXECUTION]: No remote code execution or package installation patterns were identified.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external text input for humanization, creating a potential surface for indirect prompt injection. However, this is inherent to its primary purpose.
  • Ingestion points: User-supplied text provided for stylistic rewriting.
  • Boundary markers: None; the skill does not explicitly instruct the agent to treat input as data only.
  • Capability inventory: The skill uses file manipulation tools (Read, Write, Edit, Grep, Glob) and user communication (AskUserQuestion).
  • Sanitization: No input validation or sanitization logic is present within the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 01:42 PM
Security Audit — agent-trust-hub — humanizer