obsidian-canvas-creator
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user text content to generate structured JSON data, which creates a surface for indirect prompt injection. * Ingestion points: User-provided text content (SKILL.md). * Boundary markers: The skill relies on natural language analysis to structure content but lacks formal delimiters for untrusted data. * Capability inventory: The skill is restricted to generating JSON output and does not have the ability to execute shell commands, perform network requests, or modify the file system. * Sanitization: The skill provides explicit instructions for character escaping, such as converting specific quote types and escaping double quotes, to ensure the resulting JSON is valid and safe to parse.
- [SAFE]: No malicious patterns, persistence mechanisms, or obfuscated code were detected. The included layout algorithms are purely mathematical and focused on geometric positioning.
Audit Metadata