Intelligent work tracking built around cognitive ergonomics — button-driven sessions, interval check-ins, and AI-powered summarisation that turns brief notes into valuable documentation without disrupting your flow.
The core philosophy: minimise friction, maximise documentation. SheepCat is built so you can focus on the work, not on tracking it.
Your tracking session begins and ends at the press of a button. No background scanning, no automatic triggers — you decide when your day starts and when it ends.
SheepCat runs quietly in the background between check-ins. There are no constant interruptions — just a gentle prompt at the interval you choose, so your concentration stays intact.
A brief note — a ticket number and a few words — is all you need. AI does the heavy lifting, turning that quick entry into a polished, professional update that genuinely adds value.
Minimal interruption by design — SheepCat only asks when it matters.
Prompts appear at your chosen interval — never mid-flow, never at random. When the time comes, a gentle notification asks what you've been working on.
A simple dialog asks for a ticket number and a brief description. No lengthy forms — just the essentials.
Enter a short note like "fixed login bug" and let AI expand it into a professional, detailed log entry — so you stay focused on coding.
Currently powered by Ollama for fully local, private AI — with Anthropic Claude and OpenAI planned for future releases.
SheepCat runs AI entirely on your machine using Ollama (default model: deepseek-r1:8b). Fast, private, free, and fully offline — no account, no API key, no cost.
Anthropic Claude and OpenAI are planned as future provider options for users who prefer cloud-based summarisation. The modular design makes integration straightforward.
After each check-in interval, AI generates a concise summary of the tasks completed — ready for standups or personal review.
At session end, all interval summaries are consolidated into a comprehensive daily report covering key accomplishments.
AI automatically groups work by ticket number so your logs are structured and easy to reference — without manual organisation.
Each log entry captures system information — operating system, machine name — providing full context alongside your work notes.
Adapt SheepCat to your workflow through a simple JSON configuration file.
Currently uses Ollama for local, private AI processing. Support for additional providers — such as Anthropic Claude and OpenAI — is planned for future releases.
Set your preferred check-in frequency. The default is 60 minutes, but you can choose any interval that matches your rhythm.
Customise where CSV log files are saved and how they are named — ideal for integrating with your existing folder structure or tooling.
Configure how long SheepCat waits for an AI response before timing out — keeps the app responsive on slower machines.
Control which summary types are generated — hourly, daily, or both — and where they are written.
A local configuration file lets you override any setting without touching the shared config — perfect for teams with different preferences.
Every minute of your day is captured — in a portable, open CSV format you always own.
Start and end timestamps with duration tracking are recorded for every log entry, giving you an accurate picture of how your time was spent.
Each entry is tagged with a ticket number for easy project and billing reference — whether you use Jira, GitHub, or your own system.
Both your original notes and the AI-generated summaries are saved side by side, so you always have the raw input and the polished output.
OS and machine name are captured automatically, making it straightforward to filter logs if you work across multiple devices.
Mark entries as resolved with a single click or double-click toggle — so you can track which tickets have been completed.
All data is stored as plain CSV files. Open in Excel, import into databases, or process with your own scripts — no proprietary format lock-in.
Instantly find any past work log entry — then let AI surface patterns and insights across your history.
Search across yesterday, the last 7 days, or any custom date range. Quickly zero in on exactly the period you need.
Type any keyword to find matching entries across your entire work history — useful for digging up old ticket references or project notes.
Ask the LLM to analyse your matched entries — uncovering themes, patterns, and a progress summary from your own work data.
Export search results — with or without AI analysis — as a formatted Markdown report, ready to share with your team or save for later.
Every matching entry shows its timestamp, title, and AI summary side by side so you have complete context at a glance.
Search is performed locally — no network required, no delays. Results appear instantly from your CSV log files.
Stay focused with a built-in task list that lives alongside your work log — no context-switching to another app.
Create to-do items with High, Medium, or Low priority so you always know what to tackle next.
Set tasks to repeat daily or on specific days of the week — great for recurring standups, reports, or personal routines.
Double-click any task to instantly toggle it between Pending and Done — the fastest way to check off what you've completed.
Archive completed tasks to keep your list clean without permanently deleting your history.
Select multiple tasks and mark them all as Done or Pending in one action — perfect for sprint reviews or end-of-day wrap-ups.
A status bar shows total items and pending count so you can see your workload at a glance without scrolling through the list.
Developers rarely work on just one thing — SheepCat keeps up with you.
Log work against multiple ticket numbers in the same session — context-switching is tracked automatically without losing time against any ticket.
Time is tracked and attributed per ticket, giving you an accurate breakdown of how long was spent on each item.
Hourly summaries group all active tickets together so you get a clear, unified view of your hour — regardless of how many tasks you juggled.
Push your AI-generated work summaries directly to your project management tools — always on your terms, never automatically.
Post AI-generated work summaries as comments on Jira issues — keeping your tickets updated without switching tabs or copy-pasting manually.
Push updates to Azure DevOps work items in the same way — one click from right inside SheepCat, without interrupting your workflow.
Nothing is ever sent automatically. Updates only leave SheepCat when you explicitly press the send button — you are always in complete control of what goes where and when.
Preview the update before it is posted. Edit, trim, or adjust the AI-generated content so it reads exactly as you want — no surprises, no accidental posts.
API tokens and credentials for Jira and Azure DevOps are stored only on your machine — never shared with any external service beyond your own integrations.
Integration follows the same SheepCat approach — minimal friction, maximum control. No mandatory fields, no forced workflows, and easy to disconnect whenever you choose.
Fine-tune SheepCat to match your workflow — right inside the app, no config file editing required.
Switch between AI providers — currently Ollama for local, private processing, with Anthropic Claude and OpenAI planned — directly from the settings page.
Set your preferred check-in frequency from the UI. Default is 60 minutes — adjust to any interval that suits your rhythm.
Choose which Ollama model to use for summarisation (e.g. deepseek-r1:8b) and configure API endpoints without editing JSON manually.
Customise where your CSV work logs are saved — ideal for integrating with your existing folder structure or cloud sync.
Configure how long SheepCat waits for an AI response before timing out, keeping the app responsive on any hardware.
A local settings file lets you override any shared setting without touching the defaults — perfect for teams with different preferences.
First time? SheepCat walks you through setup step by step — no documentation required.
A friendly first-run wizard greets new users and explains what SheepCat does — removing the learning curve from day one.
The onboarding wizard checks whether Ollama is installed and running, with clear instructions if it isn't.
Choose your check-in interval and preferred AI model during setup — you're ready to track from the very first session.
Contextual hints throughout the interface help you discover features without needing to read an external manual.
When using Ollama, zero data ever leaves your machine — not a single byte.
Ollama runs the AI model locally on your hardware. Work descriptions, ticket numbers, and summaries never touch an external server.
Once installed, SheepCat runs entirely offline. No third-party service, cloud subscription, or internet connection is required for daily use.
Local AI processing means no per-token fees, no usage caps, and no surprise bills — just unlimited summarisation at the cost of your own hardware.
Built with well-established, cross-platform technologies for reliability and portability.
The entire application is built in Python 3.7+ — cross-platform, readable, and easy to extend or customise.
System tray integration and all UI pages are built with Tkinter/ttk, giving a native look and feel on Windows, macOS, and Linux without extra dependencies.
During an active session, pynput monitors for idle periods — useful for detecting breaks so you can keep your logged time accurate.
Plain CSV files keep data accessible and portable — no database server required, and no lock-in to a proprietary format.
A clean abstraction layer makes it easy to swap or add AI providers — Ollama today, Claude or OpenAI when you need them.
All settings live in a human-readable JSON file with support for local overrides, making configuration straightforward for any skill level.
Data access uses a repository abstraction — making it trivial to migrate from CSV to SQL, NoSQL, or a REST API as your needs grow.
A centralised theme module defines fonts, colours, and widget styles consistently across every page — ensuring a polished, cohesive look.
SheepCat fits into your workflow — not the other way around.
Automatically track all development work throughout the day with minimal manual input.
AI-generated summaries serve as living project progress documentation, built up automatically each day.
Review daily and hourly summaries to understand work patterns, blockers, and achievements over time.
Professional AI-generated summaries are ready to share with clients or stakeholders — no post-processing required.
Yesterday's daily summary gives you instant standup material — what you did, what's in progress, and any blockers.
Push AI-generated work summaries directly to Jira or Azure DevOps work items — keeping your tickets accurate with one click, whenever you're ready.
SheepCat is free for personal use — support development or get in touch about a commercial licence.
SheepCat is an open-source project built for the neurodivergent community. If you find it useful, consider supporting to help keep it growing!