AI & Agents

A local-first CLI code agent powered by Gemma on Ollama. Implements a ReAct loop with tool-calling so it can read, write, and refactor code on your machine with no cloud calls and no vendor lock-in.

A Claude Code plugin that narrates every edit the agent makes in plain English, so you can approve or undo with full context. Born out of distrust of silent diffs — human-in-the-loop without the friction.

A collection of writing prompts and Claude skills modelled on the voices of famous engineers and researchers. Works with Claude Code, ChatGPT, Gemini, and any LLM — a testbed for how far prompt-shaped style transfer can go.

An open-source Model Context Protocol server that automates full-stack CRUD scaffolding for MERN apps — from Mongoose models to React components — with a preview-before-apply workflow so the agent stays on a leash.

A fast, private, serverless background removal tool running entirely in the browser via ONNX models. No uploads, no tracking — exploring how much real ML can ship to the client in 2026.

Training a spam classifier, then systematically breaking it with perturbed inputs — a hands-on walkthrough of adversarial examples in NLP.

GPT-4o's multimodal capabilities — native training across text, voice, vision, and real-time latency improvements.

Text-to-image mechanics — diffusion models, text encoding, iterative denoising, and current limitations.

Express.js meets OpenAI API — building a personality-driven chatbot with system prompt engineering.