fal-restore

The fal-restore skill is an Open Design catalogue entry advertising an image restoration and quality-fixing workflow built on fal.ai's hosted restoration models. According to its SKILL.md, it covers deblurring, denoising, face fixing, and restoring old documents. It addresses the problem of recovering usable quality from degraded, blurry, noisy, or aged images without needing local restoration tooling. Trigger phrases include "fal restore", "restore image", "deblur", "denoise", "fix faces", and "document restore".

This SKILL.md is a lightweight pointer curated by the fal.ai community team rather than a full implementation. It names the capability and lists triggers, but the original assets, scripts, and references live upstream in the fal-ai-community/skills repository. To run the complete workflow, a user installs the upstream bundle into their active agent's skills directory and invokes the skill by name or trigger phrase. Using it in practice involves fal.ai's hosted models.

The intended users are photographers, archivists, restoration hobbyists, and anyone digitizing or cleaning up old photos and documents, along with marketplace operators cataloguing image-generation and enhancement capabilities. Those who need the actual restoration logic should follow the upstream link.