joelhooks

joelhooks / joelclaw

Public

Personal AI operating system — blog, architecture decisions, and the journey from zero to a composable agent system.

joelclaw.com
43
3
100% credibility
Found Feb 17, 2026 at 18 stars 2x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

JoelClaw is a personal website documenting the journey, decisions, discoveries, and hardware for building a custom always-on AI assistant inspired by OpenClaw.

How It Works

1
🌐 Discover JoelClaw

You stumble upon joelclaw.com while reading about personal AI setups and click to visit.

2
📝 Read the latest writing

Browse short articles and notes about building a custom AI helper from scratch.

3
🔍 Explore cool discoveries

Check out interesting tools, ideas, and projects that agents automatically captured and summarized for you.

4
📋 Review key decisions

Dive into the list of architecture choices explaining exactly how and why the system was built this way.

5
🖥️ See the hardware setup

View the network of computers powering the always-on AI, like the main Mac Mini hub.

💡 Get inspired

Walk away understanding a real personal AI system inside and out, ready to experiment with your own.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 43 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

Overview

JoelClaw documents building a personal AI operating system on a Mac Mini, pulling from OpenClaw patterns. TypeScript monorepo with a Next.js site for blog posts, ADRs, agent-captured "Cool Finds," and network status. Backend runs event-driven agents via Inngest on k3s (Inngest, Qdrant, Redis StatefulSets).

What It Does

Serves a clean personal GitHub pages-style site with: - Blog on the build journey. - Filterable ADRs tracking decisions like adopting Inngest for agent loops or Redis-backed state. - Agent-summarized discoveries (repos/tools/articles) from casual mentions. - Cluster overview (M4 Pro hub + storage nodes).

Agents handle pipelines: video download/transcribe/summarize to Vault, session observation to vector store, autonomous coding loops (plan/test/implement/review/judge via Claude/Pi). CLI (`igs`) sends events, checks runs, manages loops.

Syncs Vault content (ADRs/discoveries) to site automatically.

Standout Features

- Durable agent loops implement PRD stories with TDD, retries, retrospectives. - Memory pipeline distills Claude sessions into Qdrant facts/narratives. - Personal operating manual template via ADRs—proposed/accepted/superseded statuses. - Examples of personal operating system: event bus for logs, content enrichment, discovery capture.

Who It's For

Devs prototyping personal GitHub Copilot setups or agent systems. Ideal if you want a personal GitHub repository for work agents, personal operating principles in code, or personal operating model (McKinsey-style decisions as ADRs). Skip if seeking production-ready libs.

Verdict

Inspiring blueprint for personal operating system examples—shows composable agents in action. 17 stars and 1% credibility mean it's raw; fork as your personal GitHub account for work experiments. Strong on architecture transparency, weak on polish. Worth starring for agent loop patterns.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.