A Telegram bot that pulls from your RSS subscriptions and sends you a daily digest.

Problem

I follow 15+ blogs and news sources. I got tired of checking each site manually. One bot, one message per day.

Architecture

RSS feeds → Parser → Deduplication → Digest formatter → Telegram Bot API

Key features

• Add/remove feeds via bot commands (/add, /remove, /list)
• Daily digest at a configured time
• Mark articles as read
• Filter by keyword

Setup

pip install feedparser python-telegram-bot apscheduler
export TELEGRAM_TOKEN=your_token_here
python bot.py

Status

v1.0.0 — Running in production for 2 months. No issues.

A real-time system monitor that runs entirely in the terminal.

Why terminal?

Sometimes you’re SSHed into a machine with no GUI. A terminal dashboard gives you CPU, memory, disk, and network stats without installing any heavy monitoring tools.

Built with

• Textual — TUI framework
• psutil — system metrics
• rich — pretty tables and charts

Screenshot (ASCII)

┌── CPU ──────────────────┐  ┌── Memory ────────────────┐
│ Core 0  ████░░░░  42%   │  │ Used   6.2GB / 16GB  38% │
│ Core 1  ██░░░░░░  21%   │  │ Swap   0.1GB / 4GB    2% │
└─────────────────────────┘  └──────────────────────────┘
┌── Disk ─────────────────────────────────────────────────┐
│ /        80GB / 250GB  32%  ██████░░░░░░░░░░░░░░░░░░░░ │
│ D:\     167GB / 250GB  67%  ██████████████░░░░░░░░░░░░ │
└─────────────────────────────────────────────────────────┘

Status

v0.1.0 — functional but unstyled. Polishing in progress.

Keep your notes in Markdown. Sync them anywhere. No proprietary formats.

Problem

I take notes in Markdown across multiple machines. I wanted something simpler than Obsidian Sync — just a CLI that diffs and pushes changes.

Commands

# Initialize sync config
mnotes init --remote user@server:/notes

# Sync (two-way)
mnotes sync

# Show diff before syncing
mnotes diff

How it works

1. Hash every local .md file
2. Compare with remote hash index
3. Push new/changed files, pull remote-only files
4. Conflict = keep both with timestamp suffix

Dependencies

• paramiko — SSH
• watchdog — optional auto-sync on file change

Status

v0.2.1 — stable for two machines. Three-way sync is future work.

Classify images by description — no cloud API needed, everything runs locally.

How it works

Uses OpenAI CLIP to compute similarity between an image and a list of text labels. The highest-scoring label wins.

from PIL import Image
import clip, torch

model, preprocess = clip.load("ViT-B/32")
image = preprocess(Image.open("photo.jpg")).unsqueeze(0)
labels = ["a cat", "a dog", "a car", "a tree"]
text = clip.tokenize(labels)

with torch.no_grad():
    logits, _ = model(image, text)
    probs = logits.softmax(dim=-1)

best = labels[probs.argmax()]
print(f"Predicted: {best}")

Interface

Flask-based web UI. Drop an image, enter labels, get a classification.

If you’ve ever broken a project by upgrading a package globally, you need venvs.

What is a venv?

A virtual environment is an isolated Python installation — it has its own pip and packages, completely separate from your system Python.

Creating one

python -m venv .venv
source .venv/bin/activate    # Linux/macOS
.venv\Scripts\activate       # Windows

Installing packages

pip install requests numpy pandas
pip freeze > requirements.txt

Restoring later

python -m venv .venv
pip install -r requirements.txt

Using with conda

If you use conda, each conda create -n myenv python=3.11 works the same way — isolated by default.

Everything you need to manage conda envs day-to-day.

Create & activate

conda create -n myenv python=3.11
conda activate myenv
conda deactivate

Install packages

conda install numpy pandas matplotlib
conda install -c conda-forge some-package
pip install package-not-in-conda        # fallback

Manage environments

conda env list                          # list all envs
conda info --envs                       # same, more detail
conda remove -n myenv --all             # delete env
conda rename -n oldname newname         # rename (conda 23.x+)

Export & reproduce

# Export (conda + pip)
conda env export > environment.yml

# Reproduce on another machine
conda env create -f environment.yml

Update

conda update conda                      # update conda itself
conda update --all                      # update all packages

Quick tip

Use conda-forge as default channel for better package coverage:

A lightweight desktop application built with pywebview + Flask that wraps common Git operations into a clean visual interface.

Motivation

I found myself running the same three Git commands dozens of times a day. This tool turns those into a single click.

Features

• Stage, commit, and push in one view
• Multi-repo management
• Commit message templates
• Push history log

Tech Stack

• Backend: Python + Flask
• Frontend: HTML/CSS/JS via pywebview
• Packaging: PyInstaller

Status

v0.3.0 — stable for personal use. Multi-repo support in progress.