Image Watermarker - Antigravity Agent

Project Description

A desktop GUI application written in Python (Tkinter + Pillow) that allows users to upload a single image and automatically apply either a text or logo watermark. This eliminates the need for heavy software like Photoshop for simple branding operations (like prepping photos for Instagram). Developed iteratively with an agentic team.

Estimated Cost

Category	Description	Cost
Development Labor	Agent-driven development	Agent Cost
AI Tokens	Google Deepmind / AI Studio usage	Nominal
Production Hosting	n/a (local execution only)	$0.00
Other	Google Stitch logo creation	$0.00
TOTAL		Nominal

Mockup

The Tkinter UI will feature a strictly utilitarian design—focusing on functional layouts rather than modern stylized themes. The primary view will have an image preview area, buttons to select an image, select a watermark, and output settings (compression vs original resolution).

To Run

In Your Local Environment

(Instructions pending development configuration)

Product Roadmap (Deliverables)

MVP (Must Do)

• Build a utilitarian Tkinter window that accepts an image upload.
• Implement text watermarking onto the image using the Pillow library.
• Implement image/logo watermarking using the Pillow library.
• Implement an options toggle for output: Maintain original resolution vs Compress.
• Incorporate Google Stitch/AI Studio to design and generate the actual logo to be used as a test watermark.
• Export the final watermarked image to a local directory.
• Achieve an 80% test coverage baseline (focusing on underlying Pillow processing logic rather than brittle GUI mocks).

Stretch Goals (Should Do at some point)

• Batch processing (processing multiple images at once).
• Expand the UI to allow dynamic watermark positioning (e.g., clicking corners).
• Adjustable watermark opacity slider.

Super-Stretch Goals (Could Do at some point)

• Auto-scaling the watermark relative to the size of the uploaded image.

Out of Scope (Won't Do)

• Complex modern UI themes (customtkinter, etc).
• Cloud hosting or server-based deployments.

Design

Flowchart (Program Logic Flow)

(Pending formal technical design phase)

Development Workflow

• 1. Logo Workshop (Stitch/AI Studio)
• 2. Core Image Processing (Pillow API wrapper)
• 3. Utilitarian GUI Scaffold (Tkinter)
• 4. Connect GUI to Logic
• 5. Write Tests for pytest (Target: 80% coverage on core logic)

Reflection

DATE	COMMENTS
2026-02-23	Phase 1 & 2 Completed via Project Kickoff Workflow. Initial repository filled with standard files and PRD Lite generated.

References

• Pillow Documentation
• Tkinter Documentation