just-html

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructions explicitly prohibit the use of external JavaScript libraries, CSS resets, and icon fonts, which effectively prevents supply chain attacks and unauthorized remote script execution.
  • [DATA_EXFILTRATION]: The skill prevents data exfiltration and credential theft by explicitly forbidding network operations (fetch/API) and access to browser storage mechanisms (localStorage/sessionStorage).
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data to generate code. 1. Ingestion points: User-provided brand tokens, text content, and design references (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: Generation of a single HTML/CSS/JS code block; specifically forbids network calls, file writes, and external script loading. 4. Sanitization: Instructions do not specify input validation or escaping. The risk is assessed as safe due to the significant capability restrictions placed on the generated output.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 02:59 PM
Security Audit — agent-trust-hub — just-html