DISPATCH

Five agents. Three epics. One terminal you forgot to check. Dispatch is a zero-config browser dashboard that shows every agent's live status, phase progress, and what got left unfinished — so you always know what happened while you were looking elsewhere.

---

The problem

You have multiple AI coding agents running across terminals, coordinating through a shared .tasks/ directory. Tracking what's happening means cat-ing plan files and grepping through logs — and you only find out an agent stalled after the fact. Dispatch watches .tasks/ and shows everything live in a browser.

Who this is for

Dispatch is for anyone running AI coding agents — whether that is two Claude Code agents in two terminals, a Gemini CLI crew, or a mixed fleet of Claude, Copilot, Gemini, and Codex agents splitting a large epic. If your agents write to .tasks/, Dispatch works without any setup.

Windows requires WSL for the heartbeat wrapper. Everything else is stable. If you hit a bug, open an issue — we respond fast.

What you get

• Multi-provider agent support — unified agent management across Claude Code, GitHub Copilot CLI, Gemini CLI, and OpenAI Codex. Agents from all tools appear in a single sidebar with provider badges. Auto-discovery imports agents from any installed CLI at startup.
• Know who is alive — heartbeat-based status dots show every agent's liveness at a glance; a stalled or crashed terminal turns red immediately
• Track epic progress without opening files — phase railroad tracks show exactly where each workstream stands across every registered project
• See what got left behind — the Tasks view shows every phase that wasn't finished across all epics, sorted by what's claimable now. No more wondering if the plan actually completed.
• Follow completed work in real time — the activity feed streams phase summaries as agents finish; click any entry to read the full execution log and handoff notes
• Cross-provider session tracking — monitor active sessions and token usage across Claude, Copilot, Gemini, and Codex from a single dashboard. Provider filter narrows any view to one or more tools and persists in the URL.
• MCP server — 10 tools for any MCP-connected agent to observe state, update phases, and spawn other agents. Auto-injected into project .mcp.json and .codex/config.toml on registration. Zero manual config.
• Watch multiple projects at once — register any number of repos; Dispatch watches them all simultaneously
• Zero config — if your agents already write to a .tasks/ directory, Dispatch works out of the box
• Agent RPG system — your fleet has personality now. Each agent has a class, stats, and a progression track tied to actual output.

Agent classes

Each persona maps to an RPG class with base stats (speed, quality, thoroughness, creativity on a 1–10 scale) derived from how that role actually works. A few examples:

Persona	Class	Title
ceo	Visionweaver	Strategist
cto	Archon of Code	Architect
staff-engineer	Merge Knight	Warrior
devops	Deploymancer	Warlock
tech-writer	Lore Keeper	Bard
custom	Freelancer	Rogue

XP comes from existing AgentProfile data — tasks completed (50 XP each), epic contributions (200 XP), delegations (15 XP), sessions (10 XP), and milestone bonuses. No additional backend required. Agents progress through 50 levels across five tiers: Apprentice, Journeyman, Expert, Master, and Ascended.

For the full class table, XP formula, and tier breakdown, see Features — Agent RPG System.

For the complete feature catalog, see Features.

Quick start

1. Clone and install:

git clone https://github.com/thepixelabs/dispatch.git
cd dispatch
npm install
chmod +x start-agent.sh

2. Start the server:

npm run dev
# Development: open http://localhost:5173
# Production build (npm run build && npm start): open http://localhost:4242

3. Register a project to watch (in a new terminal):

node cli.js add /path/to/your/project --name "My Project"

4. Confirm the dashboard shows your project:

Open http://localhost:5173. Your project should appear in the project tab bar. If you have existing .tasks/ epics they will load immediately.

5. Start an agent with heartbeat tracking (optional):

./start-agent.sh ceo /path/to/your/project

The agent appears live in the dashboard. Open more terminals, start more agents, watch them work.

6. Try the sample project (optional):

Want to see the dashboard with realistic data before setting up a real project? Register the included sample project:

node cli.js add /path/to/dispatch/sample-project --name "Sample Project"

It ships with two epics, phases in various states (DONE, IN_PROGRESS, BLOCKED), and execution log entries — so the dashboard looks alive immediately. See sample-project/README.md for details.

7. Install the bundled agent personas (optional):

npm run install-agents

Documentation

The full documentation lives on the Dispatch docs site:

Document	Contents
Feature Catalog	All features by category — includes multi-provider support, MCP server tools, provider setup, cost tracking per provider
How It Works	The .tasks/ convention, multi-provider session tracking, MCP server, data flow
CLI Guide	Project management, multi-CLI agent setup, dashboard spawn
Integrations	MCP tools, per-provider setup (Claude, Codex, Copilot, Gemini)
Agent Execution	Adapter architecture, Run button, SSE streaming, session resume
Security	G1-G4 security gates, prompt sandboxing, threat model
Contributing	How to contribute, adding new provider adapters, code standards

License

Apache License 2.0. See LICENSE.

---

_{Made by @thepixelabs. Contributions welcome.}