cursor-docs-canvas

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structured guidance for generating interactive documentation canvases and relies on a local Python adapter for rendering when native platform features are unavailable.
  • [SAFE]: The implementation in scripts/cursor_canvas_adapters.py includes rigorous path validation logic (_resolve_input, _safe_output, _inside_root). These functions verify that all file operations remain within the designated project root and explicitly reject symlinks to prevent path traversal or unauthorized file access.
  • [SAFE]: The script implements sensitive data protection through a redaction engine (_redact_preference_text). This mechanism identifies and masks API keys, access tokens, passwords, and absolute user paths within transcript records before they are processed or rendered into artifacts.
  • [SAFE]: The documentation rendering process includes HTML sanitization and URL validation (_safe_href) to ensure that generated links use safe protocols (http, https, mailto) and that cross-references do not resolve to locations outside the project environment.
  • [SAFE]: No network operations, shell command executions, or external dependency installations were detected. The skill uses only standard Python libraries for string processing and local file I/O.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 12:15 PM