humanizer-zh
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted user-supplied text and file content, which serves as a surface for indirect prompt injection attacks.
- Ingestion points: The skill accepts user-provided text snippets and processes existing file content when in 'file mode'.
- Boundary markers: The instructions do not specify explicit delimiters to isolate untrusted user content from the agent's instructions.
- Capability inventory: The skill directs the agent to rewrite text and perform file modifications (writing back the final draft) in certain modes.
- Sanitization: Instructions are provided to preserve technical structures (code blocks, frontmatter, data) during the rewriting process, offering structural integrity but not preventing instructional injection.
- [EXTERNAL_DOWNLOADS]: The project documentation references installation via a GitHub repository.
- Evidence: The README suggests installing the skill from a GitHub repository using the 'npx skills' command.
Audit Metadata