After years deep in enterprise stacks—writing C#, optimizing SQL queries, and wrestling with JavaScript—I realized something incredibly frustrating: the very ticketing systems I used to track my work were actively fighting against how my brain processes information.
If you have a "neurospicy" brain, you probably know the feeling. You sit down, open a massive Jira board or a nested to-do list, and immediately hit a wall of executive dysfunction.
Traditional project management tools demand high cognitive load. They require you to manually categorize, update, and track context across multiple screens. For someone dealing with ADHD or executive dysfunction, that friction is exhausting.
I needed a system that adapted to my focus, rather than forcing me to adapt to a rigid workflow. So, I stepped outside my usual stack, learned some Python, and built SheepCat-TrackingMyWork.
It’s an open-source, Dockerized app designed specifically to take a gentle, neuro-inclusive approach to task tracking. I built it to survive my own workday, and I'm sharing it in the hopes it helps someone else.
The Problem: Time Blindness and The "Wall of Awful"
The biggest hurdle with executive dysfunction isn't doing the work; it's the transition states and remembering what work you actually did. By 5:00 PM, time blindness kicks in, and daily stand-up prep becomes a panic-inducing exercise of checking git commits and browser histories just to piece the day together.
I didn't need another complex UI to manage. I needed a tool that acts like a gentle assistant tapping me on the shoulder.
The Solution: A "Gentle" Approach to Tracking
I designed SheepCat around a few core philosophies that specifically cater to neurodivergent needs:
1. Interstitial Logging (No heavy lifting)
Instead of forcing you to open a dashboard and fill out a form, SheepCat runs in the background. At customizable intervals, a simple Tkinter-based prompt gently asks: "What task have you done?" You type a single sentence, drop in a ticket reference (like DEV-405), and hit enter. It doesn't demand your attention; it just captures your current state and gets out of your way.
2. Automated Stand-up Prep via Local AI
This is where the magic happens. The app is fully containerized and hooks into a local Ollama instance (I've been using models like deepseek-r1:8b). Every hour, it generates a quick summary of your logs. At the end of the day, the LLM reads through your entire CSV history and writes a cohesive summary of your tasks and main takeaways. It pieces the puzzle together so your brain doesn't have to.
3. 100% Private and Local
Because brain-dumping tasks is highly personal, and enterprise work is confidential, sending my logs to a cloud API wasn't an option. Everything in SheepCat runs on your own hardware. Zero data leaks, zero privacy concerns.
Built to be flexible
While the current version uses a simple CSV storage model, I utilized a modular repository pattern. Whether the community wants to adapt it for SQL, NoSQL, or external APIs later, the architecture is ready for it.
I’ve open-sourced the whole thing under the GNU AGPLv3. If you struggle with time blindness, context switching, or just the sheer cognitive weight of standard enterprise tools, I invite you to try it out.
GitHub Repo: SheepCat-TrackingMyWork
Documentation: Setup & Website
(Note: For enterprise teams wanting to implement neuro-inclusive tools without the AGPL network-copyleft restrictions, I do offer a commercial license on the site).
Let's make our workflows adapt to us, not the other way around. If you try it out, I’d love to hear your feedback on the UI or how the Ollama integration handles your daily summaries!
Top comments (0)