Python Hidden Gems

100 Python mini-projects using interesting, lesser-known, and powerful packages from PyPI. The hidden gems that solve real problems elegantly but rarely get attention.

MIT Licensed 20 Working 80 WIP — Good First Issues 100 Projects

Phase 1 — Working Now

20 done
01

The Glitch Art Generator

pillow, numpy

Corrupts image channels & spatial warps to create glitch art

Working
02

Audio-Reactive Terminal

soundcard, rich

Real-time audio visualization in your terminal

Working
03

Automated YouTube Shorts

yt-dlp, moviepy

Downloads videos and extracts short clips automatically

Working
04

Facial Recognition Security Cam

face-recognition

Webcam face detection with logging and alerts

Working
05

PDF Presentation Maker

fpdf2

Generates PDF slide decks from markdown-ish input

Working
06

QR Code Art

qrcode, pillow

Generates stylized QR codes with embedded images

Working
07

Structured Output Enforcer

pydantic, instructor

Forces LLM outputs into validated JSON schemas

Working
08

Local Vector Brain

chromadb

In-memory vector search without a server

Working
09

Codebase Interrogator

radon, ast

Analyzes Python codebases for complexity & maintainability

Working
10

The Aggressive Linter AI

ruff

Runs ruff with custom rules and AI-suggested fixes

Working
11

Self-Healing Web Scraper

httpx, tenacity, selectolax

Scrapes with retry, fallback, and auto-repair

Working
12

Local Audio Transcriber

faster-whisper

Offline speech-to-text using Whisper models

Working
13

Billion-Row Cruncher

polars

Processes 1M+ rows in seconds using Polars

Working
14

The Ultimate JSON Parser

orjson, msgspec

Parses & validates JSON 10x faster than stdlib

Working
15

Async Network Scanner

asyncio

Concurrent port scanner using raw asyncio

Working
16

Log Anomaly Detector

glom, regex

Extracts & detects anomalies in structured logs

Working
17

Memory Leak Hunter

pympler, tracemalloc

Tracks Python memory allocations & finds leaks

Working
18

The Immutable Cache

diskcache

Persistent, fast, file-backed cache that survives restarts

Working
19

Resilient Data Miner

httpx, parsel

Mines data from multiple sources with XPath resilience

Working
20

Dark Web RSS Feed

feedparser, httpx

Aggregates RSS feeds from any source with filtering

Working

Phase 2 — Agentic AI & Data Engineering

21-40
21

API Wrapper Generator

datamodel-code-generator

Auto-generates Python API clients from OpenAPI specs

WIP
22

The God Mode Shell

prompt_toolkit

AI-enhanced shell with command suggestions and history search

WIP
23

Folder Guardian

watchdog

Monitors directories and triggers actions on file changes

WIP
24

Interactive DB Explorer

textual

TUI for browsing SQLite/Postgres with rich rendering

WIP
25

Universal Converter CLI

pint, pandas

Converts between units, currencies, and data formats

WIP
31

Agentic Knowledge Graph Builder

networkx, spacy

Builds knowledge graphs from unstructured text

WIP
34

Multi-Agent Debate Arbitrator

autogen

Multiple AI agents debate and reach consensus

WIP
35

LLM Cost Token Router

litellm

Routes LLM calls by cost, latency, and token budget

WIP
37

Semantic Codebase Search

chromadb, tree-sitter

Search code by meaning, not just text

WIP
40

Synthetic Test Data Factory

faker, mimesis

Generates realistic test data for any schema

WIP

Phase 3 — DevOps, Infrastructure & Security

41-80
62

Zero-Day Vulnerability Scanner

pip-audit, safety

Scans dependencies for known CVEs

WIP
65

Agentic Cloud Cost Optimizer

boto3, rich

Analyzes cloud spend and suggests savings

WIP
68

Self-Healing Docker Orchestrator

docker, tenacity

Auto-restarts unhealthy containers with diagnostics

WIP
73

Zombie Process Hunter

psutil

Finds and kills orphaned processes consuming resources

WIP
76

Agentic Prompt Optimizer

dspy

Automatically optimizes LLM prompts for better outputs

WIP
78

RAG Hallucination Detector

guardrails-ai

Detects when RAG systems hallucinate beyond sources

WIP

Phase 4 — Media, Multimodal & Generative

81-100
86

Autonomous Video Highlight Editor

opencv, moviepy

AI detects and extracts highlights from long videos

WIP
89

Multimodal Document Parser

unstructured, pytesseract

Parses PDFs with text, tables, images, and charts

WIP
92

Autonomous Meme Generator

pillow, openai

AI generates memes from text prompts

WIP
95

Autonomous Slide Deck Generator

python-pptx, openai

AI creates presentation slides from a topic

WIP
98

Audio Deepfake Detector

librosa, sklearn

Detects AI-generated audio with spectral analysis

WIP
100

Universal Media Transmuter

ffmpeg-python

Converts any media format to any other format

WIP

Python Hidden Gems by Annapurna Agentic Solutions — MIT Licensed

View on GitHub