ux-writing
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze interface copy from untrusted external sources, such as Figma designs or user-provided text strings, to provide UX writing feedback.
- Ingestion points: External copy is provided by the user or fetched via Figma MCP links as described in SKILL.md and docs/claude-figma-integration.md.
- Boundary markers: The instructions do not define clear delimiters or 'ignore instructions' warnings to separate untrusted copy from the agent's core instructions.
- Capability inventory: The skill is focused on text auditing and generation; it does not utilize dangerous subprocesses, file-write operations to sensitive paths, or unauthorized network tools.
- Sanitization: There are no specific instructions for filtering or sanitizing the content of the analyzed copy.
Audit Metadata