image-enhance · Image Enhancement

image-enhance

Image quality enhancement: high-quality upscaling (Lanczos 2x/4x) plus denoising, sharpening, and automatic contrast/saturation adjustment for blurry, dark, or noisy images.

Run a one-click --auto pass or control each step — denoise, sharpen, contrast, saturation — individually. Default is 2x upscale; upscale can also be disabled to enhance in place. Outputs the enhanced image plus an execution report (original vs. new dimensions, which enhancements were applied).

Example invocation: "Upscale this old photo 2x and denoise it" (with the image attached).

Full brief

Positioning

image-enhance is a deterministic image-quality tool: high-quality upscaling, denoising, sharpening, and tone adjustment for mildly blurry, noisy, or dark images. It is not AI super-resolution and cannot invent detail — it improves, not redraws.

Core capabilities

Workflow

  1. Inspect the image and identify the problem (blur / noise / darkness)
  2. Pick parameters: add --denoise for noisy shots; --scale 1 --auto for tone-only fixes
  3. Run skills/shared/scripts/img_enhance.py enhance; outputs land in outputs/<topic>/

Inputs & outputs

Input Notes
Image Required — the image to enhance
Upscale factor Optional — 2 by default; 1 = enhance without resizing
Enhancement steps Optional — one-click --auto, or denoise/sharpen/contrast/saturation individually

Output: enhanced image plus execution report, both under outputs/<topic>/.

Boundaries with adjacent skills

Skill Lane
image-enhance (this) Traditional quality improvement: upscale, denoise, sharpen, retone
ai-image-gen AI image-to-image upscaling/redraw that can generate new detail
image-editing Routine operations: resize, crop, compress, watermark — no quality improvement

Fit

Before you start


image-enhance is part of the Aiglade Skill library. Invoke it from the Aiglade chat box in plain language.

← All 155 skills