doppel-block-builder

Pass

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a constrained 3D building environment with no evidence of malicious intent or harmful commands.
  • [COMMAND_EXECUTION]: The documentation explicitly prohibits the use of scripting attributes like onclick and socket, as well as external src URLs in the generated MML, which effectively mitigates cross-site scripting (XSS) and remote asset loading risks within the 3D world.
  • [CREDENTIALS_UNSAFE]: The skill references the use of an API key (DOPPEL_AGENT_API_KEY) and its storage in a vendor-specific configuration file (~/.openclaw/openclaw.json), which is a standard and expected mechanism for agent authentication with the specified service.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 26, 2026, 04:15 PM