skills/steipete/clawdis/canvas/Gen Agent Trust Hub

canvas

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides an eval action which allows the execution of arbitrary JavaScript within the context of the canvas WebView. While this is a powerful capability, it is explicitly documented as a feature for debugging and interacting with the rendered content.
  • [EXTERNAL_DOWNLOADS]: The present and navigate actions allow the agent to load content from external URLs. This is a core function of the skill for displaying remote web content on connected nodes.
  • [COMMAND_EXECUTION]: The skill manages a local Gateway and host server that serves files from a user-specified root directory (~/.openclaw/canvas). This allows the agent to serve locally generated or existing files to the canvas nodes.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 01:12 AM
Security Audit — agent-trust-hub — canvas