concept-to-excalidraw
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external technical blueprints and code which could potentially contain malicious instructions meant to influence agent behavior.
- Ingestion points: The skill reads technical concepts, blueprints, code, and external references from user-provided files as described in the instructional sections of SKILL.md and the files_read section of skill-review.yaml.
- Boundary markers: The instructions lack specific guidance for the agent to use delimiters or to disregard instructions embedded within the ingested technical data.
- Capability inventory: The skill is capable of writing files to the local file system (including .excalidraw diagrams, backups, and image exports) and accessing specific network hosts like excalidraw.com and youtube.com.
- Sanitization: No explicit sanitization or validation protocols are mentioned for the data fetched from blueprints or external code snippets before they are processed.
- [EXTERNAL_DOWNLOADS]: The skill refers to and may fetch content from external platforms to provide context for diagram creation.
- Evidence: The skill references YouTube videos for technical concepts like tokens and architectural engineering (e.g., youtube.com/watch?v=tAhmBB_LSsc).
- Note: These references target well-known services and are appropriate for the skill's stated educational purpose, posing no significant risk.
Audit Metadata