studioai-canvas

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were found in the provided files. The skill focuses on its primary purpose of canvas management.
  • [COMMAND_EXECUTION]: The skill interacts with defined MCP tools for canvas operations. It does not contain any arbitrary shell command execution or unauthorized access patterns.
  • [DATA_EXFILTRATION]: The skill manages project data through specific API calls to the Convex backend. No exfiltration of sensitive files, credentials, or unauthorized network transmission was detected.
  • [PROMPT_INJECTION]: Instructions are focused on technical task execution (adding/wiring nodes) and do not include techniques to override agent behavior or bypass safety policies.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 12:15 PM
Security Audit — agent-trust-hub — studioai-canvas