Photoshop CS6 Automation Driver

The Problem

Photoshop CS6 still handles many production graphics tasks well, but repetitive work can remain manual: posters, white-background removal, watermarks, batch resizing, and ID-photo background changes are easy to get wrong when parameters or files are handled by hand. Scripting CS6 also has sharp edges because CS6 APIs differ from newer versions. Some features are not exposed through the DOM, some require AM events, and filters, masks, smart objects, and layer styles have different limits.

How It Works

The skill maps a natural-language request into a local CS6 execution path rather than emitting generic code. It first extracts canvas size, text, colors, image paths, and export targets, then looks up task recipes. It prefers documented DOM APIs when they are clean and usable; when the DOM is missing or too coarse, it falls back to vetted AM snippets. If local Photoshop is unavailable, it produces a manually runnable .jsx script and tells the user to open it through Photoshop's script runner, instead of failing hard.

Boundaries

It is aimed at Windows environments where CS6 is installed locally, with exports focused on JPEG, PNG, and PSD. It is not a replacement for newer AI editing features: there is no generative fill, neural filters, subject selection, or sky replacement, and direct WebP / HEIF read/write is not supported. Liquify, hair-level refinement, and other complex retouch tasks may only get approximate scripted options or a reminder to use manual work. Destructive edits should default to duplicating the background layer, and saving over files or removing layers needs explicit confirmation.