Hi HN,
I built SheepCat-TrackingMyWork originally to solve my own daily struggles with staying organized and remembering what I actually achieved by 5 PM.
I've spent my career heavily focused on C#, SQL, and JavaScript, using traditional, rigid enterprise ticketing systems. The problem is those tools clash with how my brain processes tasks. Dealing with executive dysfunction and context switching meant I needed a tool that actively helps me track my day without requiring me to constantly manage a complex UI. I built this primarily for myself, but I'm sharing it in the hopes that it might help other "neurospicy" folks who find standard tools overwhelming.
How it works:
Instead of a giant, static to-do list, the app runs in the background and gently prompts you every hour: "What task have you done?" You can also manually add tasks in the meantime.
If connected to a local Ollama instance, the app takes those entries, generates an hourly summary, and saves everything to a local CSV. At the end of the day, the local LLM reads through your daily logs and generates a full summary of all tasks and main takeaways. It basically handles your time-tracking and daily stand-up prep automatically.
The Tech & Privacy:
I stepped out of my usual stack and built the core application in Python. Because work logs are highly personal, keeping data private was a strict requirement. The app is containerized with Docker and hooks into an external Ollama setup to run the LLM locally. Not a single byte of data goes to a third-party API.
The License:
I’ve open-sourced it under the GNU AGPLv3 so anyone can use, modify, and self-host it for free. However, I am also offering a paid commercial license for enterprise teams looking to implement neuro-inclusive tools without the network-copyleft restrictions of the AGPL.
I’m currently transitioning from just building this for myself to being an open-source maintainer, and I'd love your feedback:
Does the hourly prompt + CSV logging sound like a good balance between actively tracking work and minimizing cognitive load?
Thoughts on the Python/Docker architecture and how it manages the external Ollama hooks?
For those who have gone the AGPL/Commercial dual-license route, what were your biggest hurdles converting enterprise users?
Website/Docs: https://chadders13.github.io/SheepCat-TrackingMyWork-Website...
Happy to answer any questions about the code, the Ollama integration, or the design choices!
So I've been working on a online tool that makes it simple to practice basic maths. Its in its early stages and I'm plan to keep adding more questions types and content too it.
So currently have count, simple whole sum, powers and percentages. Some of them have settings to adjust the difficulty too so it can be made easier or harder.
There are currently some simple graphs to show your questions results and will keep improving on them as the site develops.
I want to make it easier for anyone to be able to practice and improve their mental maths skills. It's certainly helped me while I've been working on it
