How it works
The life of a project, told in eight scenes.
One project goes from a name on a screen to something your customers rely on — and never leaves PlushLab on the way. Plush and the street cats we've taken in since will walk you through it.
01A new project
It starts with a name.
Type one and you get a project with everything it needs from day one — a roadmap, issues, docs, meetings and monitoring — at its own address.
02The code
Tie it to the code.
Install the GitHub app and pick the repositories this project uses. From then on, the work and the code keep track of each other.
03The plan
Lay out the roadmap.
Milestones with dates, epics that hold the story, issues in an ordered backlog. Turn on sprints and estimates only when your team wants them.
04The work
Build it, branch by branch.
Mention PL-128 in a branch or pull request and the issue follows along — branch, review, merge. Merged means done, and nobody has to move a card.
05AI integration
Your AI, on the team.
Paste what ChatGPT or Claude wrote — a spec, a table, a whole HTML prototype — and it keeps its shape, safely sandboxed. Or let your assistant work in the project through the API: read issues, add specs, report progress.
06Deploy
Ship it to your own server.
PlushLab builds your Dockerfile, pushes the image and rolls it out on your server — migrations first, the new container checked before the old one stops. One click rolls back. Every deploy knows its commit, its pull request and the issues it closes.
07Logs · traces · metrics · profiling
See inside every request.
Point your OpenTelemetry exporter at the project and logs, traces, metrics and CPU profiles land in one place — with deploys drawn on the charts, so you see what changed. A new error becomes an issue in one click.
08The proof
Sleep well. It's watched.
Healthchecks on everything you run, alerts that don't cry wolf and open an incident when something is really wrong, and a status page your customers can read — so you can actually rest.
