Flagship product · Live

ReplayCore: Minecraft gameplay, rebuilt in the browser.

ReplayCore is our own live platform and the clearest proof of what ForgeVector builds. It spans a Java developer API, a cloud backend on SQL and Cloudflare R2, a real-time browser renderer and a secure, licensed dashboard, capturing Minecraft gameplay server-side and replaying it at 1:1 fidelity.

What it is

A full platform, not a single plugin.

Most Minecraft tools stop at the server. ReplayCore goes the whole way: it captures gameplay through a developer API, moves it into the cloud, stores it efficiently, and rebuilds it in the browser so teams can review exactly what happened. It is the kind of large, integrated system we love to build.

  • Server-side capture via a Java developer API, tuned for low overhead.
  • Cloud backend that processes and indexes captures away from live play.
  • 1:1 browser rendering that reconstructs the real scene, not a video.
  • Secure dashboard with role-based access and per-customer licensing.
Product
ReplayCore by ForgeVector
Capture
Java developer API (Java Edition servers)
Backend
Cloud processing & indexing
Storage
Cloudflare R2 + SQL database
Front-end
Real-time 1:1 browser renderer
Access
Secure dashboard, role-based
Licensing
Proprietary, per customer
Status
Live now
Architecture

From the server tick to your browser.

Five stages take a moment of gameplay from a live Minecraft server all the way to a scrubbable, 3D replay in your browser.

Capture

A lightweight Java plugin records gameplay server-side through the ReplayCore developer API: world, blocks, entities and movement, tuned to stay light on the server tick.

Ingest & process

Captured data is streamed to the cloud backend, where it is processed, validated and indexed away from the live server, so capture never slows down play.

Store

Large recording payloads go to Cloudflare R2 object storage, while structured metadata, accounts and licences live in a SQL database, fast to query, cheap to keep.

Render

The browser viewer streams the data back and reconstructs the scene in real time at 1:1 fidelity: the same world, entities and movement, scrubbable like video but fully 3D.

Review & act

Teams open replays in a secure, role-based dashboard, share links and make decisions with full context, instead of relying on screenshots and memory.

Under the hood

The components that make it work.

Java developer API

A clean, documented API and plugin that records gameplay server-side with minimal overhead, and lets other developers integrate capture into their own systems.

Cloud backend

Ingests captures, processes and validates them, and exposes the secure APIs the dashboard and viewer rely on, all kept off the live game server.

SQL database

Structured data, accounts, replay metadata and licences, kept in a relational database that stays fast to query as volumes grow.

Cloudflare R2 storage

Large recording payloads live in R2 object storage, keeping heavy data durable and cost-effective while staying quick to stream back to the viewer.

Real-time browser renderer

The front-end streams capture data and reconstructs the scene in 3D, in the browser, at 1:1 fidelity, scrubbable and shareable like video but fully interactive.

Dashboard & licensing

A secure web dashboard with role-based access, on top of a per-customer licensing system that the backend validates to control access.

1:1 fidelity

A replay, not a recording.

The hardest and most important part of ReplayCore is the renderer. Instead of saving a video, it captures the underlying state of the game and rebuilds it in the browser: the same world geometry, the same entities, the same positions and movement, tick by tick. You can move through it in 3D, scrub time, and see exactly what happened from any angle.

  • Faithful reconstruction of world, blocks and entities, not an approximation.
  • Runs in the browser, no downloads, no game client needed to watch.
  • Scrub, pause and explore in 3D, with full context around every moment.
Want one built for you?

This is the kind of system we build.

ReplayCore exists because we like the hard problems: real-time capture, cloud pipelines, storage at scale and rich browser front-ends. If you need a platform like this, on top of a game or any other system, that is exactly what we do for clients.

ReplayCore is live.
Visit replaycore.com for the product, or join our Discord for setup and support.
FAQ

ReplayCore, explained.

What is ReplayCore, in one sentence?
ReplayCore captures Minecraft gameplay server-side and replays it in your browser at 1:1 fidelity, backed by a cloud platform with a developer API, SQL, object storage and a secure dashboard.
What does “1:1 fidelity” actually mean?
It means the browser viewer reconstructs the real scene, the same blocks, entities, positions and movement captured on the server, rather than showing a recorded video or an approximation. You are looking at the actual gameplay, rebuilt in 3D in the browser.
How is gameplay captured?
Through a Java developer API and plugin that runs on the Minecraft server. It records the data ReplayCore needs efficiently, with a strong focus on keeping overhead low so the live server stays smooth.
Where is the data stored?
Large recording payloads are stored in Cloudflare R2 object storage, while structured data such as metadata, accounts and licensing is kept in a SQL database. This keeps the heavy data cheap to store and the structured data fast to query.
How does licensing work?
ReplayCore is proprietary software, licensed per customer. Licences are validated by the backend, so access to capture and the dashboard is controlled and secure.
Can ForgeVector build something like this for me?
Yes. ReplayCore is the proof of what we do: developer APIs, real-time pipelines, cloud backends, storage and browser front-ends. If you need a custom platform on top of a game or another system, that is exactly our wheelhouse.

Need a platform like ReplayCore?

Real-time capture, cloud backends, storage and browser rendering are exactly what we build. Tell us what you have in mind.