humanizer-en
Pass
Audited by Gen Agent Trust Hub on Jul 14, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill's design involves processing arbitrary user-supplied text or reference files, which introduces a potential surface for indirect prompt injection.\n
- Ingestion points: Processes text input from the user or reads files from provided paths to perform style analysis (
SKILL.md).\n - Boundary markers: The instructions do not define explicit delimiters to separate untrusted user content from the agent's system instructions.\n
- Capability inventory: The skill uses
Read,Write, andEdittools to interact with the file system (SKILL.md).\n - Sanitization: Input content is not sanitized or validated before processing.\n
- Mitigation: The skill includes a 'Safety Tier' (Tier M) that instructs the agent to show results and wait for explicit user approval before overwriting any files.
Audit Metadata