Move your local dev to Flisk. We help you migrate.

Bring your docker-compose.yml, your Tiltfile or your setup wiki. We work with your team until one flisk up runs the whole stack.

  • Production stays where it is. Flisk only changes how your team runs the stack on their machines.
  • Nothing to unwind later. Flisk writes standard Dockerfiles, Vite configs and Tiltfiles into your repo. No lock-in.
  • Core is MIT licensed. Migrating needs no license key and no paid plan.

Solo project or side project? Start with the quickstart and flisk import, or ask your coding agent to migrate it.

Request migration help

We reply by email. No newsletter, no sales sequence.

Where teams move from

Each starting point has a different first step. All of them end at one manifest per service.

docker-compose.yml

Docker Compose

flisk import reads the services in your root docker-compose.yml and writes a service.json for each one. Run it with --dry-run first to see the plan without writing anything. Try it on the compose example.

Tiltfile

Tilt

Flisk runs on Tilt and generates the Tiltfile from your manifests, so live reload works the way your team already knows. We help move custom rules into service.json.

skaffold.yaml, devspace.yaml, garden.yml

Skaffold, DevSpace, Garden

These run your dev loop against a cluster. Flisk runs it on Docker on the laptop and leaves the cluster to your Helm or GitOps setup. See the comparison.

README.md, setup.sh

Scripts and a setup wiki

We start from what a new hire does on day one and turn each step into a manifest or a flisk doctor check. Then the wiki can go.

What a migration looks like

The same commands whether you do it alone or with us.

  1. Read the plan

    Import with --dry-run. It prints every service it found and writes nothing.

  2. Write the manifests

    Run the import for real, then review each service.json like any other pull request.

  3. Check every machine

    flisk doctor checks Docker, ports and services on each laptop before anyone starts the stack.

  4. Start the stack

    flisk up generates the configs and starts the services with live reload. Keep the old setup next to it until the team stops using it.

terminal
# 1. see what would be imported
flisk import ./legacy --dry-run

# 2. write a service.json per compose service
flisk import ./legacy --yes

# 3. check the machine
flisk doctor

# 4. run the whole stack
flisk up

Or ask your coding agent to migrate it

Connect Claude Code, Cursor or Codex to Flisk, open your repo and paste the prompt. The agent runs the same steps through the flisk mcp server.

terminal, in your repo
# install the Flisk CLI
npm install -g @flisk-sh/flisk

# connect your agent (pick one)
flisk connect claude
flisk connect cursor
flisk connect codex

This adds the flisk mcp server and the Flisk skills to your repo. Other MCP clients and setup details: Flisk for AI agents.

prompt
Migrate this repo's local development to Flisk.
Import the services from docker-compose.yml, show me
the plan before writing anything, then start the
stack and fix anything that fails to come up.
  1. doctorChecks that Docker and Tilt are ready on this machine.
  2. project_importRuns with dryRun first and shows you the plan, then writes a service.json per compose service.
  3. config_verifyChecks every manifest and the generated files.
  4. up, statusStarts the stack and waits until every service is ready.
  5. logsIf a service fails, reads the error, fixes the manifest or code and starts again.

The exact steps depend on your agent and your prompt. Review the service.json files it writes like any other pull request. Stuck halfway? Request migration help.

Migration help FAQs

What we do, what we don't, and what it costs.

Who qualifies for migration help?
Teams whose app runs as several services (APIs, workers, frontends, databases, queues) and who want every developer to start it the same way. If one person works on one service, the quickstart is faster than a call.
What do I need to do?
Send the form. We reply by email to ask what your stack looks like today and how your team starts it. Have your compose file, Tiltfile or setup notes at hand. Not sure Flisk fits? Read does Flisk fit first.
What does it cost?
Migration help costs nothing, and Flisk core is free and MIT licensed. Premium is a separate license for extras like Playwright and C4 diagrams at $19 per developer per month. You don't need it to migrate. See pricing.
Can you do the migration for me?
We work through it with your team. flisk import does the mechanical part and we help with what it can't infer: build contexts, ports, seed data and migrator resources. We don't ask for access to production or to your cloud accounts.
Do we have to change production or CI?
No. Flisk is for local development. Production stays on your Kubernetes, Helm or hosting setup, and CI keeps running what it runs today. Flisk with Helm shows how the two sit side by side.
What requests don't qualify?
Moving production hosting, writing cluster manifests, and single-service projects. Flisk doesn't host or deploy anything, so we can't help move a production workload.
What if we stop using Flisk?
Your own Dockerfiles, service.json files and code are never touched. flisk eject writes down where the generated files are and how to run the stack with plain tilt up, without the Flisk CLI.

Read before you move

Field notes on moving an existing codebase without stopping work.

Ready to start?

Try the import on your own repo, or send the form and we'll work through it with you.

Start free Request migration help