social-media-dashboard
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill features a surface for indirect prompt injection due to its requirement to process external design files.
- Ingestion points: The workflow (SKILL.md, Step 1) instructs the agent to read an active
DESIGN.mdfile. - Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the ingested content.
- Capability inventory: The skill is limited to generating UI code (HTML/CSS/JS) for artifact previews and does not have access to sensitive tools, file-system writing, or network exfiltration capabilities.
- Sanitization: No sanitization or validation of the
DESIGN.mdcontent is performed before interpolation. - [EXTERNAL_DOWNLOADS]: The skill generates code that references assets from well-known and trusted services.
- Evidence: The example implementation (example.html) fetches typography from Google Fonts and user avatars from GitHub's official avatar domain. These references are standard for web prototypes and do not represent a security threat.
Audit Metadata