create-canvas-extension

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: Executes repository-specific scripts including npm run plugin:validate, npm run build, and npm run website:data to verify the scaffolded extension. These commands are standard for the development lifecycle of the target repository.\n- [PROMPT_INJECTION]: Ingests user-provided metadata (name, description, author) to populate plugin.json manifests. This represents a standard indirect prompt injection surface for scaffolding tools.\n
  • Ingestion points: User input for Extension ID and display metadata requested during execution (SKILL.md).\n
  • Boundary markers: No specific delimiters are applied to user input within the manifest templates.\n
  • Capability inventory: File system write operations and execution of npm build/validation scripts (SKILL.md).\n
  • Sanitization: Relies on the user for valid inputs and the npm run plugin:validate script for schema verification.\n- [EXTERNAL_DOWNLOADS]: References a plugin schema from agent-plugins.org. This is a standard resource for manifest validation and does not involve the execution of untrusted external code.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 12:11 PM
Security Audit — agent-trust-hub — create-canvas-extension