soia-cwork-processon-diagrams

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill incorporates extensive security-by-design features:
  • Isolated Browser Automation: Uses a dedicated, isolated Playwright profile for all interactions with ProcessOn, ensuring the user's primary browser credentials and history are never accessed or impacted.
  • Strict Action Validation: Declarative browser actions are validated against a restricted allowlist. The runner explicitly forbids remote mutations (e.g., deletion, sharing) and sensitive data fields (e.g., cookies, tokens) in its interaction logic.
  • Built-in Secret Scanning: Automatically scans downloaded diagram content (VSDX and POS formats) for potential plaintext credentials (passwords, AWS signatures). Artifacts containing detected secrets are quarantined and blocked from final delivery until sanitized.
  • Detailed Auditing: Employs a multi-stage checkpoint and audit system that verifies the integrity of the inventory and download process using SHA-256 hashes.
  • [COMMAND_EXECUTION]: Internal orchestration relies on subprocess.run to chain the skill's specialized Python components (batching, state management, finalization). Execution is restricted to the skill's own script environment using the current Python interpreter.
  • [EXTERNAL_DOWNLOADS]: Interacts with official ProcessOn domains (processon.com) for diagram retrieval. All automated browser navigation is strictly restricted to HTTPS ProcessOn URLs through validation filters.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 08:45 AM
Security Audit — agent-trust-hub — soia-cwork-processon-diagrams