Describe the app once, then run it locally or render cluster YAML

[patrickleet]

What you write

You list the app’s pieces in Rust (modules, commands, models, projections). Then you say how to cut them into processes.

let module = Module::new("todo")
    .command_definitions([/* todo.create, … */])
    .models([/* TodoView */])
    .projections([/* project_todos listens to todo.created */])
    .build()?;

let manifest = Application::new("todo-app")
    .module(module)
    .build()?
    .manifest()
    .clone();

// one process that does everything
let local = compile_deployment_plan(
    "local",
    &manifest,
    [ProcessIntent::with_preset("full", &manifest, ProcessPreset::Full)?],
)?;

// or split write vs project
let split = compile_deployment_plan(
    "split",
    &manifest,
    [
        ProcessIntent::with_preset("writer", &manifest, ProcessPreset::Writer)?,
        ProcessIntent::with_preset("projector", &manifest, ProcessPreset::Projector)?,
    ],
)?;

That’s the whole description. You do not write “if postgres, connect, migrate, spawn outbox, attach GraphQL to Service…”.

What you run

Start it (same as the demo app now):

# app binary still just calls the framework
DATABASE_URL=sqlite:./app.db BIND=127.0.0.1:8791 cargo run

Under the hood that is:

let (manifest, plan) = e2e_local_plan();
let realized = RuntimeHost::realize_local(&plan, "full", options, MyApp).await?;
// realize_local: connect DB, workers, dispatcher
// then serve GraphQL

Turn the same plan into deploy files:

distributed deployment validate --manifest manifest.json --plan plan.json
distributed deployment render --manifest manifest.json --plan plan.json --target kubernetes --out out/k8s
distributed deployment render --manifest manifest.json --plan plan.json --target knative --out out/knative
distributed deployment render --manifest manifest.json --plan plan.json --target hops-xr --out out/hops

Hops YAML is real structured mounts, not debug text:

kind: DistributedApplication
spec:
  processes:
    - name: 'writer'
      mounts:
        - kind: command
          id: 'todo.create'
    - name: 'projector'
      mounts:
        - kind: projector
          id: 'project_todos'

Check contracts:

distributed contracts check --catalog distributed.contracts.json

Before → after

	Before	After
Author writes	Domain plus a custom host: SQLite vs Postgres branches, migrations, Service, outbox/consumer loops, GraphQL wired to Service	Domain + a short Application / DeploymentPlan. Framework starts the process
Local run	Copy-paste host.rs per app	RuntimeHost::realize_local(plan, …)
Cluster YAML	Missing, or Hops output like Projector { id: "project_todos" }	Same plan → k8s / knative / hops-xr. Tests parse the files
Illegal topology	Easy to ship an Atomic command without its projection	Plan compile fails before anything starts
CLI	dctl still showed up	Only distributed

Why that’s better

• Next app doesn’t reimplement the runtime.
• Local process and cluster YAML can’t silently disagree.
• Bad splits fail at compile/plan time, not in production.

Not in this PR

distributed dev (watch Cargo + Vite), live cluster apply, a new Hops package repo.

Workspace version on this branch is 0.0.0-dev. Publish tags (v4, …) still own the released crate version.

[coderabbitai]

Warning

Review limit reached

@patrickleet, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 35 minutes

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: d0ebfbfa-e20c-415c-a90d-6e41272f000d

📥 Commits

Reviewing files that changed from the base of the PR and between b97f96a and 364b04f.

📒 Files selected for processing (36)

• Cargo.toml
• distributed_cli/Cargo.toml
• distributed_cli/src/cli.rs
• distributed_cli/src/client_compiler/manifest/mod.rs
• distributed_cli/src/client_compiler/manifest/parse.rs
• distributed_cli/src/client_compiler/mod.rs
• distributed_cli/src/contracts/tests.rs
• distributed_cli/src/lib.rs
• distributed_cli/tests/cli_deployment.rs
• src/application/local_sql.rs
• src/application/mod.rs
• src/application/mount.rs
• src/application/plan.rs
• src/application/render.rs
• src/application/resolve.rs
• src/application/runtime_host.rs
• src/command_dispatch/local.rs
• src/graphql/client_manifest/codec.rs
• src/graphql/client_manifest/mod.rs
• src/graphql/http.rs
• src/graphql/mod.rs
• src/graphql/schema.rs
• src/graphql/sdl.rs
• src/graphql/surface/application.rs
• src/graphql/surface/mod.rs
• src/lib.rs
• tests/application_plans.rs
• tests/e2e-ui/crates/service/src/application.rs
• tests/e2e-ui/crates/service/src/host.rs
• tests/e2e-ui/crates/service/src/lib.rs
• tests/e2e-ui/crates/service/src/modules/contracts.rs
• tests/e2e-ui/crates/service/src/modules/graphql.rs
• tests/e2e-ui/crates/service/src/modules/mod.rs
• tests/e2e-ui/crates/service/src/oidc_layer.rs
• tests/e2e-ui/ui/src/lib/walkthrough/demos.ts
• tests/e2e-ui/ui/vite.config.ts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[patrickleet]

Closing. We should not generate Kubernetes/Knative manifests from this layer, and the hops-xr target was a fake XR (no XRD, no reconcile).

The goal is to make the existing Service startup simpler, not add a second host/plan/render model.