obsidian-canvas-creator
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user content (articles, notes, text) to generate a structured JSON canvas.
- Ingestion points: Text content provided by the user in the prompt (as described in SKILL.md).
- Boundary markers: No explicit delimiters or boundary markers are defined to separate untrusted user input from the agent's instructions.
- Capability inventory: The agent is instructed to generate a valid JSON object which is then returned to the user for use in Obsidian.
- Sanitization: Basic JSON quote escaping and specific character transformations for Chinese quotes are defined in SKILL.md, but these do not prevent high-level instruction following from within the processed data.
Audit Metadata