Overview

Holodeck gives every builder — and every agent — a numbered, isolated preview environment (a slot) that spins up a full vertical stack on demand. Reserve a slot, deploy any service at a commit, branch, or PR, exercise the real product end-to-end, then release it when you're done.

The problem it solves

Today everyone shares one dev environment, so we constantly step on each other: your testing collides with someone else's, and the moment anyone breaks shared dev, everyone who depends on it is blocked. Holodeck replaces that with your own isolated environment — you're never waiting on shared dev, and a broken shared environment no longer stops you from testing your change.

What you get

Value	What it means for you
10+ parallel environments	Every PR, workstream, or agent gets a clean slot (holodeck-1…holodeck-N) — no more waiting on shared dev/staging.
Full-stack & cross-repo	A change touching 3–8 repos runs together as one coherent environment. You test the real integration, not a mock.
Isolated data	Each slot has its own workload namespaces, databases/Firestore, Temporal namespaces, and identities — nothing leaks between slots.
Fast & ephemeral	Spins up in seconds, lives for a bounded TTL, and tears down cleanly — no lingering cost or config drift.
Built for agents	The whole lifecycle (reserve → deploy → poll → diagnose → release) is a plain HTTP/MCP API, so a coding agent can run it with no human in the loop.
Shift verification left	See your change running end-to-end before merge — catch integration problems early instead of after they hit shared envs.

In one line: Holodeck turns "wait for a shared environment and hope it works after merge" into "every change gets its own full-stack preview, on demand, in seconds."

How you reach it

• Web console: holo-server.abridge.coffee (Google sign-in / IAP).
• API: https://holo-api.abridge.coffee (internal — for agents, scripts, and CI). Requires tsh vnet or VPN.
• MCP: dev-holo-mcp-teleport for read tools right inside Cursor / Claude / Codex.

New here? Start with the Getting started guide, then read how deploys and the architecture work under the hood.