Skip to content
View joeynyc's full-sized avatar
💭
Having Fun!
💭
Having Fun!
149 followers · 58 following

Achievements

Achievement: Starstruckx3Achievement: Pull Sharkx2Achievement: QuickdrawAchievement: Pair Extraordinairex2Achievement: Public SponsorAchievement: YOLO

Achievements

Achievement: Starstruckx3Achievement: Pull Sharkx2Achievement: QuickdrawAchievement: Pair Extraordinairex2Achievement: Public SponsorAchievement: YOLO

Highlights

  • Pro

Block or report joeynyc

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
joeynyc/README.md

Hi, I'm Joey

I build local AI systems, agent tools, and interfaces.

Self-taught builder in NYC. No CS degree, no bootcamp—just a lot of building, debugging, and shipping in public. My work spans multi-DGX inference, AI agent infrastructure, native macOS apps, and local-first tools.

Featured projects · DGX Spark lab · More projects · X / Twitter

Featured projects

  • Grok UI — A local-first command center for Grok Build with runtime control, session history, and Git visibility. TypeScript Local-first
  • Honeycomb Lab — A hex-map control plane for a home AI fleet.
  • Hermes HUDUI — A real-time web dashboard for watching Hermes think, use memory, make mistakes, and grow. Python Web UI
  • Hermes HUD — A terminal consciousness monitor for Hermes Agent, built with Textual. Python TUI
  • Hermes Skins — Custom visual themes for the Hermes CLI agent. YAML Themes
  • SkillScore — A CLI that evaluates SKILL.md-based agent skills and produces actionable quality scores. TypeScript Bun

DGX Spark lab

I experiment with large local models, multi-node inference, and agent workloads on NVIDIA DGX Spark.

  • spark-doctor — A diagnostic CLI for DGX Spark that catches power, memory, thermal, CUDA, Docker, and inference-stack issues. Python CLI
  • Hy3 295B on 2× DGX Spark — Tencent Hy3 295B MoE with TP2 over 200GbE, 256K context, and agent-ready benchmarks. NVFP4 vLLM
  • Nemotron Puzzle 75B on 1× DGX Spark — A tuned NVFP4 setup with MTP, prefix caching, concurrency benchmarks, and reproducible recipes. NVFP4 Inference
  • Hermes Local MoA — Run Hermes Agent as a mixture of agents across local OpenAI-compatible model servers. Python MoA
  • Cosmos + Locate Anything — A generated-video and live-grounding pipeline for DGX Spark. Python Computer Vision
  • Hermes Desktop for DGX Spark — Connect Hermes Agent Desktop on Mac to local models through an OpenAI-compatible API. Shell Local AI

More things I've shipped

  • National Vehicle Inventory CLI — Agent-first search for real U.S. dealer inventory, with JSON, MCP, caching, and price history.
  • OpenClaw Mission Control — A native SwiftUI command center for OpenClaw agents.
  • PinDrop — An offline-first infinite canvas for sticky notes.
  • specimen — A local-first SVG gallery, inspector, and exporter in one dependency-free HTML file.
  • Govee MCP — Control Govee smart lights from AI assistants using natural language.
  • repobrief — Generate compact codebase context maps for AI coding agents.

Find me on X: @aijoey

Pinned Loading

  1. hermes-hudui hermes-hudui Public

    Web UI consciousness monitor for Hermes — the AI agent with persistent memory

    Python 1.8k 222

  2. hermes-hud hermes-hud Public

    TUI consciousness monitor for Hermes

    Python 879 64

  3. hermes-skins hermes-skins Public

    Custom skins (visual themes) for the Hermes CLI agent

    Python 528 32

  4. NousResearch/hermes-agent NousResearch/hermes-agent Public

    The agent that grows with you

    Python 221k 42.1k

  5. spark-doctor spark-doctor Public

    Local diagnostic CLI for NVIDIA DGX Spark (GB10). Detects power caps, UMA pressure, thermal risk, CUDA 13/SM_121 wheel mismatches, Docker issues, and validates vLLM/Ollama/llama.cpp/SGLang recipes.

    Python 92 7

  6. skillscore skillscore Public

    CLI tool that evaluates AI agent skills and produces quality scores. Works with any SKILL.md-based skill from skills.sh, ClawHub, GitHub, or local directories.

    TypeScript 16 2