# Cloud Control Plane Deployment

This guide covers the deployable Cloud API and outbound Device Host Agent.
Run commands from the repository root after synchronizing the uv workspace:

```bash
uv sync --locked --all-packages
```

## Local SQLite

SQLite is intended for local development and tests with one Cloud API process.
Configure public and host credentials even in local mode because anonymous
development access cannot authorize a host identity.

```powershell
$env:CLOUD_ENVIRONMENT = "local"
$env:CLOUD_DATABASE_URL = "sqlite:///cloud/cloud.sqlite3"
$env:CLOUD_PUBLIC_CREDENTIALS_JSON = '[{"principal_id":"local-sdk","token":"replace-public-token","scopes":["tasks:submit","tasks:read","pool:read","plugins:read","plugins:admin"]}]'
$env:CLOUD_HOST_CREDENTIALS_JSON = '[{"principal_id":"local-host","token":"replace-host-token","scopes":[],"host_id":"host-local"}]'
uv run --package device-cloud-api device-cloud-api --host 127.0.0.1 --port 8001
```

In a second terminal, start the Host Agent with the matching host identity and
token:

```powershell
$env:HOST_AGENT_CONTROL_PLANE_URL = "http://127.0.0.1:8001"
$env:HOST_AGENT_HOST_ID = "host-local"
$env:HOST_AGENT_TOKEN = "replace-host-token"
uv run --package device-host-agent device-host-agent
```

At startup, the Host Agent loads device registrations from
`tasks/device_config.sqlite3`, the same `DeviceConfigStore` used by the local
Runtime console API. Register or update devices before starting the Host Agent,
then restart it to reload changes. In Compose,
`HOST_AGENT_TASKS_PATH` selects the host directory mounted at `/app/tasks`; it
defaults to `./tasks`.

The Host Agent only initiates outbound HTTP requests. It does not expose an
inbound port.

## Managed Edge Enrollment

New edge installations do not need a pre-coordinated Host or device ID. The
Cloud API accepts configured one-time enrollment credentials:

```powershell
$env:CLOUD_ENROLLMENT_TOKENS_JSON = '[{"principal_id":"edge-installer","token":"replace-with-a-long-random-one-time-token"}]'
```

On the edge Host, omit `HOST_AGENT_HOST_ID` and `HOST_AGENT_TOKEN` and provide
the enrollment token only for the first successful enrollment:

```bash
export HOST_AGENT_CONTROL_PLANE_URL="https://cloud.example.com"
export HOST_AGENT_ENROLLMENT_TOKEN="replace-with-a-long-random-one-time-token"
export HOST_AGENT_IDENTITY_PATH="tasks/host_identity.json"
export HOST_AGENT_DISPLAY_NAME="Edge Mac 01"
uv run --package device-host-agent device-host-agent
```

Before its first request the Host Agent creates `HOST_AGENT_IDENTITY_PATH` with
an instance identifier and long-lived random Host secret. The cloud consumes
the enrollment token, assigns `host_id`, stores only credential digests, and
returns the assigned ID. The Host Agent then enrolls each record from
`tasks/device_config.sqlite3`, stores its cloud-assigned `device_id` in that
database, connects the resulting devices, and starts heartbeat/claim loops.

Keep the identity file and device configuration database on persistent edge
storage with permissions limited to the service account. The identity file is
a bearer secret: do not put it in an image, repository, log, or general backup.
After successful enrollment, remove `HOST_AGENT_ENROLLMENT_TOKEN` from the edge
environment. An intact identity file is sufficient for restart; if only a
device mapping is lost, device enrollment reconstructs the same cloud ID.

Enrollment tokens are one-time even when they remain in Cloud API environment
configuration: their consumed digest is stored in the database. Reusing a token
for another edge instance returns a conflict. Create a distinct token for every
edge installation.

Explicit `HOST_AGENT_HOST_ID` plus `HOST_AGENT_TOKEN` takes precedence and keeps
the previous legacy behavior, including locally selected device IDs. This is
the rollback and staged-migration path for existing deployments.

## PostgreSQL Deployment

Start from `.env.example`, replace every `change-me-*` value, and keep the
resulting `.env` file outside version control. The Compose stack contains
PostgreSQL, the Cloud API, and one Host Agent:

```bash
docker compose up --build -d
docker compose ps
```

The Cloud API container waits for PostgreSQL, runs the committed Alembic
migrations, then starts the API on port `8001`. Readiness is available at:

```text
GET http://127.0.0.1:8001/health/ready
```

## Deploying A Jenkins-Built Image

`docker compose up --build` above builds the image from source on the target
host. For environments that pull a pre-built image instead, `Jenkinsfile` runs
the pytest suite (`-m "not integration"`), builds the image from the same
`Dockerfile`, smoke-tests both console scripts (`device-cloud-api --help`,
`device-host-agent --help`), and pushes `<REGISTRY>/<IMAGE_NAME>:<BUILD_NUMBER>-<git short sha>`
plus `:latest` to the configured registry.

`compose.deploy.yaml` is the same three-service stack as `compose.yaml`
except `cloud-api` and `host-agent` reference `image:` instead of `build:`.
Set `REGISTRY`, `IMAGE_NAME`, and `IMAGE_TAG` (see `.env.example`) to the tag
Jenkins published, then deploy without a local build step:

```bash
docker compose -f compose.deploy.yaml pull
docker compose -f compose.deploy.yaml up -d
docker compose -f compose.deploy.yaml ps
```

For a deployment that manages processes outside Compose, apply migrations
before starting the new Cloud API version:

```powershell
$env:CLOUD_DATABASE_URL = "postgresql+psycopg://USER:PASSWORD@HOST:5432/device_cloud"
uv run alembic -c packages/cloud-platform/cloud/migrations/alembic.ini upgrade head
uv run --package device-cloud-api device-cloud-api --host 0.0.0.0 --port 8001
```

Production startup requires `CLOUD_ENVIRONMENT=production`, a current schema,
and at least one configured bearer credential.

## Credentials And Scopes

`CLOUD_PUBLIC_CREDENTIALS_JSON` is a JSON array of public API principals. Grant
only the scopes required by each integration:

- `tasks:submit`: submit tasks.
- `tasks:read`: read task status and failure metadata.
- `pool:read`: list hosts and devices.
- `plugins:read`: list installed plugin registrations.
- `plugins:admin`: register installed plugin entry points.

`CLOUD_HOST_CREDENTIALS_JSON` contains Host Agent principals. Every entry must
include exactly one `host_id`; its token is valid only for heartbeat, claim,
renewal, and result operations for that host.

`CLOUD_ENROLLMENT_TOKENS_JSON` contains bootstrap principals with only
`principal_id` and `token`. These credentials cannot submit tasks, read the
pool, or operate as a Host; they can only create one durable Host binding.
Use high-entropy values generated by the deployment secret manager.

Do not place bearer tokens in command history, image layers, Compose files, or
logs. Use environment injection or the deployment platform's secret manager.
Rotate a token by deploying the updated Cloud API credential set and Host Agent
configuration together.

Dynamically enrolled Host credentials are stored as digests in the cloud
database. This release exposes repository-level revocation rather than a public
administration endpoint. An operator with database deployment access can revoke
a Host without deleting its task history:

```bash
export HOST_ID="host-..."
uv run --package device-cloud-platform python - <<'PY'
import os
from cloud.database import CloudDatabase
from core.models import utc_now

database = CloudDatabase(os.environ["CLOUD_DATABASE_URL"], create_schema=False)
try:
    changed = database.repository.revoke_enrolled_host(
        os.environ["HOST_ID"],
        revoked_at=utc_now(),
    )
    print("revoked" if changed else "not an enrolled host")
finally:
    database.close()
PY
```

## Cloud Console (Web UI)

The repository ships an independent Vue 3 + Vite SPA at `cloud-console/` that
renders the task queue/history, device pool, host registry, and plugin
registry, and exposes the existing plugin-registration action. It authenticates
the same way `CloudClient` does: by attaching a pre-issued bearer token to
every request. There is no login or session system.

### Provision an operator bearer token

Add a `CLOUD_PUBLIC_CREDENTIALS_JSON` entry whose scopes cover what the
console operators need to do. The least-privilege set for read-only dashboards
is `tasks:read`, `pool:read`, and `plugins:read`. Add `tasks:submit` only if
operators should submit ad-hoc tasks from the same tab, and `plugins:admin`
only if operators should register plugins:

```json
[
  {
    "principal_id": "console-operator",
    "token": "replace-with-a-long-random-opaque-token",
    "scopes": ["tasks:read", "pool:read", "plugins:read", "plugins:admin"]
  }
]
```

Rotate the token the same way as any other credential entry: deploy the
updated Cloud API credential set and instruct operators to paste the new token
into the console. The console keeps the token only in browser `sessionStorage`
for that tab; closing the tab discards it.

### Configure the CORS allow-list

The Cloud API has no CORS middleware by default. Before a browser can call it
cross-origin, set `CLOUD_CONSOLE_CORS_ORIGINS` to a comma-separated allow-list
that includes the exact origin (scheme + host + port, no trailing slash) the
operator's browser will load the console from:

```bash
# Allow a local Vite dev server
export CLOUD_CONSOLE_CORS_ORIGINS="http://127.0.0.1:5173"
# Or a deployed origin
export CLOUD_CONSOLE_CORS_ORIGINS="https://console.example.com"
```

Restart the Cloud API after changing this env. The middleware is added only
when the allow-list is non-empty — existing deployments see no behavior change
until an operator opts in. Blanket `allow_origins=["*"]` is intentionally not
supported because every console request carries a bearer token.

### Run the console

```bash
cd cloud-console
cp .env.example .env.local
# Edit .env.local if your Cloud API is not at http://127.0.0.1:8001
npm install
npm run dev
```

Vite prints a local URL (default `http://127.0.0.1:5173`). That exact origin
must be in `CLOUD_CONSOLE_CORS_ORIGINS` on the Cloud API. Open the dev URL,
paste the operator token, and the dashboards become available.

For a production build, run `npm run build` and serve the resulting `dist/`
behind any static file server or CDN, with `VITE_CLOUD_API_BASE_URL` baked in
at build time. The deployed origin must be in `CLOUD_CONSOLE_CORS_ORIGINS`.

### Same-origin deployment (baked into the Cloud API image)

The Jenkins-built Docker image already carries the SPA at `/app/console-static`,
and `compose.yaml` / `compose.deploy.yaml` set
`CLOUD_CONSOLE_STATIC_DIR=/app/console-static` on the `cloud-api` service. In
this mode the Cloud API itself serves the console at `/console/` (visiting `/`
307-redirects there), so operators can open `https://<cloud-api-host>:8001/`
directly — no separate dev server, no static host, no CORS allow-list needed
(the SPA and the API share one origin).

The SPA shell (`index.html`, JS, CSS) is served without a bearer token by
design — `_authorize(...)` is called inside the `/v1/*` route handlers, not in
middleware, so the SPA can boot before the operator pastes a token. All
`/v1/*` API calls still require `tasks:read`/`pool:read`/`plugins:read` scopes
as before.

To opt out (e.g. for local development where you run `npm run dev`), leave
`CLOUD_CONSOLE_STATIC_DIR` unset. The mount is conditional on that env var.

Jenkins build args (`NODE_IMAGE`, `NPM_REGISTRY`, `UV_IMAGE`, `APT_MIRROR`,
`UV_INDEX_URL`) default to CN mirrors so builds don't time out pulling from
Docker Hub / ghcr.io / npmjs.org / deb.debian.org. Blank any of them to fall
back to the upstream.

## Runtime AI Planner

The Host Agent reuses the local Runtime planner. AI planning is disabled by
default. Configure it in the Host Agent environment when goal assignments must
use a model:

```text
AI_PLANNER_ENABLED=true
AI_PLANNER_PROVIDER=anthropic
AI_PLANNER_MODEL=claude-sonnet-5
AI_PLANNER_TIMEOUT_SECONDS=30
ANTHROPIC_API_KEY=<secret manager reference>
```

For OpenAI, set `AI_PLANNER_PROVIDER=openai`, choose the deployed model through
`AI_PLANNER_MODEL`, and provide `OPENAI_API_KEY`. Provider credentials belong
only on the Host Agent; the Cloud API does not need them.

## Operational Limitations

Run exactly one scheduler-enabled Cloud API process. SQLite supports only the
documented single-control-plane development mode. PostgreSQL row locking makes
assignment and claim transactions safe if requests overlap, but this release
does not implement scheduler leader election or claim active-active scheduler
operation. Starting multiple Cloud API replicas would start one scheduler loop
per replica and is outside the supported deployment topology.

Device execution provides at-least-once side-effect semantics, not exactly-once
semantics. A device action can succeed immediately before the Host Agent loses
its lease or its result response, after which the control plane may retry the
task. Lease renewal and cooperative stop checks prevent later interruptible
actions where possible, but they cannot roll back an action already sent to a
device or safely terminate an in-progress synchronous driver call.

Use bounded attempts, inspect task attempt and failure metadata, and design
device workflows to tolerate repeated actions when the target operation allows
it. Do not use this release for operations that require a transactional
exactly-once guarantee across the cloud database and an external device.

## Shutdown And Rollback

For a normal shutdown, stop Host Agents first so they stop polling, interrupt
later cooperative actions, finish terminal reporting where the lease remains
valid, and attempt a final heartbeat. Stop the Cloud API after Host Agents have
exited, then stop PostgreSQL only if the database itself is being maintained:

```bash
docker compose stop host-agent
docker compose stop cloud-api
docker compose stop postgres
```

For rollback:

1. Stop all Host Agents and the Cloud API.
2. Back up PostgreSQL or the SQLite database file.
3. Before rolling back to a release without enrollment support, provision
   temporary static Host credentials for every managed edge that must continue
   operating. Stop those Host Agents and set their explicit Host ID/token.
4. If the previous application version cannot use the current schema, run the
   tested downgrade while no application process is connected:

   ```bash
   uv run alembic -c packages/cloud-platform/cloud/migrations/alembic.ini downgrade -1
   ```

5. Restore the previous application image or checkout and start the Cloud API.
6. Verify `/health/ready`, then restart Host Agents with credentials compatible
   with the restored Cloud API.

Downgrading revision 0002 removes dynamic credential bindings and durable
device enrollment mappings. It retains the revision-0001 Host heartbeat rows,
pooled devices, queued tasks, attempts, and plugins.

Do not remove the PostgreSQL volume during an application rollback. Queued and
attempt history are durable database state and should remain available to the
restored or forward-deployed control plane.