I need to clean AI-related parameters from files generated by tools like Stable Diffusion or DALL·E so prompts aren’t leaked.

This feature removes embedded AI-generation parameters (such as prompts, negative prompts, model details, and other generation metadata) from image and media files. It helps prevent accidental disclosure of sensitive prompts or workflow details when sharing or publishing AI-generated content.

This feature is designed to sanitize files produced by AI image generation tools by stripping AI-related parameters that can be stored inside file metadata. It targets information that may reveal how an image was created, such as prompts, negative prompts, seeds, sampler settings, model identifiers, and tool-specific generation parameters. Users can run the cleaner before sharing images publicly to avoid leaking proprietary prompts, brand guidelines, or internal creative workflows. It supports common scenarios where AI tools embed data into PNG text chunks, EXIF/XMP fields, or other metadata blocks that can be extracted by third parties. The feature reduces the risk of prompt theft, unintentional disclosure of sensitive terms, or exposure of internal experimentation settings. It is useful for creators publishing portfolios, businesses sharing marketing assets, and teams distributing images to clients or external collaborators. The workflow is straightforward: select the files to sanitize, run the cleaning process, and output cleaned copies that are safe to distribute. By removing only AI-related parameters and metadata, the feature helps maintain visual fidelity while improving privacy and control. It can be incorporated into a pre-publish checklist to ensure consistent, repeatable sanitization across batches of files.