comfyui-lora-training
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The content consists entirely of instructional documentation and configuration templates for local machine learning training workflows. No executable scripts or binary files are provided within this skill file.
- [SAFE]: No sensitive file paths, credential exfiltration, or unauthorized network operations were identified. References to local directories like
{ComfyUI}/models/loras/anddataset/are appropriate for the skill's stated purpose. - [INDIRECT_PROMPT_INJECTION]: The skill involves the agent processing external datasets, specifically text captions (.txt files) associated with images. This creates a theoretical surface for indirect prompt injection if malicious instructions are embedded in the user-provided dataset.
- Ingestion points: User-provided image datasets and text caption files (.txt).
- Boundary markers: No explicit delimiter or "ignore embedded instructions" warnings are included in the provided templates.
- Capability inventory: Metadata requires the
pythonbinary, suggesting the agent may facilitate the execution of training scripts or file operations. - Sanitization: None present in the documentation file.
Audit Metadata