humanize-with-rephrasy
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill requires a
REPHRASY_API_KEYand correctly implements security best practices by instructing users to store it in environment variables. It includes explicit guardrails to prevent the key from being hardcoded or exposed in outputs. - [COMMAND_EXECUTION]: Includes a bundled Node.js script (
scripts/humanize.mjs) to communicate with the Rephrasy API. The script is restricted to its humanization purpose and implements appropriate request timeouts. - [EXTERNAL_DOWNLOADS]: Recommends the use of the
@rephrasy/mcpNPM package for model context protocol integration, which is an official resource owned by the vendor. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text for transformation. While this constitutes a data processing surface, it is the primary intended utility of the skill and uses vendor-specific infrastructure for the task.
Audit Metadata