Projects

Turn a browser task into a recorded run and a reusable script.

Manage skills across coding agents. Install, sync, and inspect what each agent uses.

Research the web, save progress, and resume interrupted work to produce sourced reports.

Keep a personal web archive with page captures, Markdown, and screenshots.

A curated catalogue of awesome agentic AI patterns

Automatically tracks and visualizes a Hacker News user's karma points over time using GitHub Actions

Build local mock environments and reproducible evaluations for AI agents from an OpenAPI document.

Official Steel browser tools for the Pi coding agent.

Distraction-Free Writing firmware for the M5stack ESP32 based Tab5 device

An opinionated collection of Agent Skills I use daily in my work.

A Community-Driven Mapping of AI Development Tools

Turn any webpage into perfect LLM input with shareable, permanent links

A lightweight web viewer for Claude Code conversation files with a terminal-inspired ASCII aesthetic

Test how AI agents use command-line tools. Inspect the steps, find the friction, and compare runs.

Database built from scratch in Zig as an educational project with benchmarks as source of truth

🚀 AI-focused dotfiles for remote VM development. Optimized for Claude Code, Amp, and modern CLI tools. One-liner install for instant productivity.

Multi-agent co-emergence experiment exploring how identities emerge through constrained communication

Autonomous task execution plugin for LLM CLI - refactored modular architecture

Peeking-safe streaming inference for A/B tests and online metrics

A standards-compliant OpenAPI validator backed by a public, versioned Torture Suite.

Scannanas play on Literature and Latte Scrivener, as Tauri app

OSpec v1.0 - Outcome-Driven AI Agent Frameworks specification and website

Static analyzer for Marp Markdown decks - validates slide quality for AI agents

Universal open standard for compute sandboxes - Protocol-first, production-ready API for E2B, Modal, Vercel, Cloudflare & more

lmdb-tui is a terminal user interface for exploring LMDB databases

Train a 4.8M parameter dual-head transformer on Google Colab to optimize nightclub admission strategies

AI-first browser automation and web scraping tools for LangChain with advanced document loading capabilities

Chrome extension using local Gemini Nano for rewriting the tone of your copy

A time-aware, schema-rewritable, WASM-powered memory store for autonomous agents

Simple storage and organization tool

LMDB MCP Server - Model Context Protocol server for Lightning Memory-Mapped Database integration

Research on how AI agents perceive and process information

Labruno is an agent coordinator that creates multiple AI solutions to your coding tasks using parallel sandboxes and evaluates them to find the best implementation.

An Open-Source benchmarking framework designed to evaluate and compare various code execution sandbox environments

A simple command-line utility to count tokens in text files or from piped input.

devcontainer.ai - AI generated standardized dev environments