humanize-text
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill is configured to download and install executable code from an external GitHub repository (
github.com/gregorymm/humanize-text) using thenpx skills addcommand during its installation process. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data such as raw text, Figma design content, and screenshots to perform its primary function of text analysis and rewriting. This creates an attack surface where instructions embedded in the processed data could attempt to influence the agent's behavior.
- Ingestion points: Raw text input, Figma URLs accessed via an MCP server, and UI screenshots.
- Boundary markers: The instructions do not define clear delimiters or use "ignore embedded instructions" warnings for the processed data.
- Capability inventory: The skill's primary capability is text analysis and rewriting; the provided files do not show additional subprocess, file-write, or network operations beyond the installation script.
- Sanitization: No sanitization, validation, or escaping of the ingested external content is described in the provided files.
Audit Metadata