excalidraw-obsidian
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data from the $ARGUMENTS field to generate file content.
- Ingestion points: The $ARGUMENTS parameter in SKILL.md receives the user's diagram description.
- Boundary markers: No explicit delimiters or instructions are used to distinguish user input from the skill's structural instructions.
- Capability inventory: The skill writes files to the local filesystem (~/notes/obsedian/Excalidraw/) and interacts with MCP tools.
- Sanitization: The input is interpolated directly into the markdown and JSON sections without validation or escaping.
Audit Metadata