gpt-taste
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill takes user input and uses it to generate designs and reports that are written to a persistent wiki directory, creating a vulnerability surface.
- Ingestion points: User prompt provided for the design task.
- Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when writing content to the wiki files.
- Capability inventory: The skill performs file writes to llmwiki/wiki/draft/uiux/, llmwiki/wiki/index.md, and llmwiki/wiki/log.md.
- Sanitization: The skill lacks instructions to sanitize or escape user-influenced content before writing to the file system.
- [EXTERNAL_DOWNLOADS]: Fetches placeholder imagery from the well-known service picsum.photos for UI demonstrations.
Audit Metadata