01Live
clawd.run
Cloud runtimes for Hermes and OpenClaw agents with persistent memory and a real computer.
ProductOpen clawd.run at clawd.run
JoePro AI / systems infrastructure and product studio
Open any of it before you email me.
17 public projects, each with a live address and real infrastructure behind it. Look first. Then bring me something to build or something that broke.
17
17
4
02Live
Followtronics
Public-audience review queues with visible evidence and confidence.
03Live
Neural Salvage
A permanent-state platform for preserving AI memory, identity, media, and durable records on Arweave.
04Live
QR Canvas
Scannable image generation and QR compositing for art, menus, product images, and more.
05Live
JoePro Press
Public tools including the Morning Brief, image generation, model racing, and analysis utilities.
06Live
Prompt Racer
A side-by-side race interface for comparing model response speed.
07Live
Xonetization
X creator payout and eligibility tracking.
08Live
StakeSmith
NFL betting blueprints and EV simulations tied to live odds.
09Live
Food Empire
A food-trading game with recipe, pantry, costing, waste, and supplier tools.
10Live
WhatYouSay
Offline voice translation for mobile devices.
11Live
Carapace
A deterministic memory-integrity firewall and trust ledger for memory-first agents.
Open sourceOpen Carapace at carapace-xi.vercel.app
12Live
Longtake
Unbroken generated video takes with spoken characters.
13Live
The Wringer
Work-order audits and evidence-backed multi-agent execution runs.
14Live
CoreFish
Crowdsourced fishing heatmaps, catch logs, and a fishing companion.
PortfolioOpen CoreFish at corefish.joepro.ai
15Live
Agent Soul
A focused encrypted archive and restore flow for an agent's memory, persona, and configuration.
16Live
Rhyme Protocol
Creator tools for lyrics, cover art, video, audio, and the systems that help artists earn more from their work.
17Live
Surgical Jiu-Jitsu
A field atlas of precision grappling with anatomy, technique chapters, and coach material.
Routes
Five outcomes. Every project sits under exactly one.
If you already know what you are trying to get done, start here instead of with the product names.
Create
You need something made.
Video takes, scannable images, and the public tool surface where the generation actually runs.
Automate
You need work done without you in the loop.
Hosted agent runtimes, audited work orders, and a way to see which model is worth the wait.
Preserve
You need state to survive the tooling around it.
Permanent encrypted archives, and a firewall that keeps agent memory honest before it is written.
Decide
You need evidence before you commit.
Payout tracking, betting blueprints, and audience review that shows its work.
Explore
You want to see where the edges are.
Games, field data, and offline translation. These are where the odd ideas get tested in public.
Also running
Recently added to the field and not yet filed under an outcome.
Two doors
Bring me something to build, or something that broke.
Build
You have an idea, a spec, or a system that does not exist yet. I scope it, build it, and hand back something running, with the source and a plain description of how it works.
Start a buildRepair
You have something live that is failing, drifting, quietly burning money, or that nobody on the team will touch any more. I find the fault, fix it, and leave it changeable.
Start a repair