GridWork HQ Documentation

GridWork HQ is a self-hosted agency operations dashboard built for solo agency owners who want AI-powered automation without giving up control. It ships as two packages — a Next.js dashboard and a Node.js pipeline server — that work together to run your entire agency from one interface.

What You Get

• Dashboard — Next.js 16 app with client tracking, pipeline management, finance overview, cron monitoring, and an AI chat assistant. Authenticated via GitHub OAuth.
• Pipeline Server — Node.js HTTP server that spawns Claude Code processes to execute 17 AI pipelines. Handles job queuing, cron scheduling, and SSE streaming back to the dashboard.
• Knowledge Vault — A git-backed repository of templates, client files, pipeline definitions, and memory files that the AI uses as context for every pipeline run.
• 12 Cron Jobs — Automated tasks for knowledge maintenance, output archiving, weekly updates, scope audits, email sync, and more.
• Document Compiler — 7 compile types that transform raw pipeline output into branded, client-ready documents.

Architecture

┌──────────────────────────┐
│   GridWork HQ Dashboard  │  ← Next.js 16 + TypeScript + Tailwind v4
│   (port 3000)            │     SQLite via Drizzle ORM
│                          │     GitHub OAuth authentication
│   - Mission Control      │
│   - Client Tracker       │
│   - Pipeline Kanban      │
│   - Finance Panel        │
│   - Cron Dashboard       │
│   - AI Chat (FAB)        │
└────────┬─────────────────┘
         │ Bearer token auth
┌────────▼─────────────────┐
│   Pipeline Server        │  ← Node.js HTTP server
│   (port 8750)            │     Claude Code spawner
│                          │     12 cron jobs
│   - Job queue (max 3)    │     SSE streaming
│   - Cron scheduler       │
│   - Telegram alerts      │
└────────┬─────────────────┘
         │
┌────────▼─────────────────┐
│   Knowledge Vault        │  ← Git repository
│                          │     Pipeline definitions
│   - system/rules/        │     Client folders
│   - clients/             │     Memory files
│   - memory/              │     Output archive
│   - .output/             │
└──────────────────────────┘

Quick Links

Getting Started

• Prerequisites — What you need before installing
• Quick Start — From ZIP to running dashboard in under 10 minutes
• First Pipeline Run — Run the prospect pipeline and see results

Installation

• Local Setup (Mac/Linux) — Full local development setup
• VPS Deployment — DigitalOcean or Hetzner self-hosted
• Vercel + Tailscale — Dashboard on Vercel, pipeline server on your machine
• Docker — Container-based deployment

Core Features

• AI Pipelines — 17 pipelines covering lead generation, client delivery, and operations
• Dashboard Modules — Every page in the dashboard explained
• Cron Jobs — 12 automated scheduled tasks
• Knowledge Vault — Templates, client folders, and memory system

Configuration

• Environment Variables — Every env var explained
• Integrations — Notion, GitHub, Stripe, Vercel, GA4, and more

Reference

• Troubleshooting — Common issues and fixes
• Upgrading — Version migration guides
• FAQ — Frequently asked questions
• Changelog — Release history

Tech Stack

Requirements at a Glance