html-artifacts-plus
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses local workspace files such as markdown, JSON, and YAML to populate interactive editors and visualizations. This behavior is properly constrained by explicit instructions that forbid network exfiltration, hidden background processes, and direct file system writes.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from local files into generated HTML artifacts. It mitigates potential risks through the use of specific boundary markers (data-bind attributes) and a round-trip workflow that mandates user review and agent validation before any patches are applied to the source files.\n- [DYNAMIC_EXECUTION]: The skill generates HTML files containing embedded JavaScript for UI logic and diff generation. The JavaScript is generated locally and uses an audited internal diffing utility, avoiding the risks associated with remote code execution or untrusted third-party script dependencies.
Audit Metadata