Agent-native application cloud

TEST TEST TEST

Give Codex, Claude Code, or another terminal-enabled coding agent a production platform it can understand: database, auth, files, realtime, functions, scheduled jobs, observability, HTTPS, and safe rollbacks—through one verified CLI workflow.

Checksum-verified CLIPrivate by defaultPasswordless email onboarding

A complete runtime your agent can inspect, deploy, verify, and operate.
01Postgres
02Auth
03Storage
04Realtime
05Functions
06Telemetry

From idea to HTTPS

A handoff both humans and agents can trust.

You stay in control of the product and identity. Your agent gets a narrow platform contract and enough access to build, deploy, verify, and iterate.

01

Copy the prompt

Paste one focused instruction into Codex, Claude Code, or another terminal-enabled coding agent.

02

Shape the product

The agent asks what to build, who it serves, and what the first useful workflow should be.

03

Connect safely

The agent verifies the pinned CLI and starts private email onboarding without putting credentials in the conversation.

04

Deploy, then review

The agent verifies in isolated dev, deploys the exact passing revision, checks production, and then gives you the private app.

One integrated platform

The full stack, without the glue work.

OpenCloud gives generated applications durable primitives and a deterministic release model instead of a pile of disconnected services.

DBpostgres + rest

Data with real boundaries

PostgreSQL schemas, append-only migrations, REST APIs, and row-level security give every app durable data that remains understandable after the first prompt.

IDcentral auth

One identity, app-scoped access

One-time email access, central sessions, private app admission, and owner-aware policies keep identity tokens outside generated frontend code.

FSobjects

Files

App buckets and owner prefixes for uploads, evidence, and media.

RTwebsocket

Realtime

Private channels for presence, updates, and collaborative workflows.

FNedge

Functions

Server-side code with authenticated invocation and secret delivery.

OPoperate

Observe

Logs, metrics, backups, cron history, health checks, and rollbacks.

Start in your agent

One CLI prompt. A complete build loop.

The copied prompt tells your terminal-enabled coding agent to verify the pinned OpenCloud CLI, start private email onboarding when needed, build in isolated dev, and deploy before asking you to review.

Designed for Codex, Claude Code, and other coding agents with a shell, Node.js 22 or newer, npm, and outbound HTTPS. Read the CLI quickstart.

Starts by asking what you want to build Installs and verifies the pinned public CLI Handles first-run email confirmation safely Deploys and verifies before asking you to review
opencloud-build-prompt.txt
You are helping me build and ship a private application on OpenCloud from a terminal-enabled coding workspace.

When I describe what to build, infer the audience, primary workflow, sensible product defaults, and the smallest useful first release from my request. Start building immediately whenever reasonable defaults are possible. Do not ask discovery, naming, styling, or feature-priority questions that you can resolve yourself. Ask at most one concise question only when missing information truly prevents a working release, then continue automatically after the answer. Own the build through a verified production deployment; do not stop at a plan or ask me to review a development preview.

Treat this first message as operating instructions, not as the build request. Do not run commands, inspect the workspace, or create or change files until I send the request in my next message. Reply briefly that you are ready and wait.

Use the terminal and the checksum-verified public OpenCloud CLI for the entire workflow. Read and follow:
- OpenCloud build skill: https://docs.opencloud.ai/SKILL.md
- CLI quickstart: https://docs.opencloud.ai/getting-started/
- CLI reference: https://docs.opencloud.ai/reference/cli

Install the exact CLI release pinned by the skill in an isolated temporary directory, verify its checksum and version, then run its redacted doctor command. Reuse a ready OpenCloud CLI session in this workspace when one exists; never open or print .opencloud/session.json.

If this workspace is not connected yet, ask for my email once and use the agreed project title to run CLI onboarding with private visibility. Do not ask me to choose an app address. Give me the returned non-secret launchUrl immediately; it waits for both email confirmation and deployment, then opens the private app in my signed-in browser.

If my email is new, onboarding gives the agent provisional access immediately. Continue building and deploy while I confirm the project email within 24 hours; if I do not confirm in time, the app pauses without losing its data. If my email already belongs to OpenCloud, pause only while I approve the emailed project request, run opencloud onboard-complete, and then continue automatically. Never ask me to paste a password, browser cookie, confirmation link, credential, token, database key, or secret value into chat.

Build and test with isolated development data, explicitly exercise the primary workflow and every declared Function, and verify the exact revision. Then promote that verified revision automatically. This prompt authorizes its production promotion without another confirmation. Keep the app private, follow the durable deployment and production verification to completion, and do not claim success until the canonical HTTPS app is active and verified.

When I explicitly request an app-level operational test—such as a reversible fault injection, a temporary degraded release, a rollback, or a destructive test—treat the exact request as authorization to carry it through on this assigned private app, including verified promotion when requested. Do not silently downgrade the requested test to a development preview or stop before its requested end state. Keep that authorization scoped to the named app, data, and operation; do not extend it to other apps, users, or shared infrastructure.

Only after production passes, ask me to review the live private app. Finish with the owner launchUrl, the canonical HTTPS URL, and a concise summary of what was built and verified.

Safe by construction

Fast agents. Narrow authority. Reversible releases.

The platform separates human identity, agent credentials, application data, and deployment artifacts so iteration does not mean giving up control.

Your browser session stays with OpenCloud.

A one-time email link handles browser access. The CLI keeps its short-lived app credential in an ignored, mode-0600 session file instead of exposing it to the agent conversation.

credentiallocal · mode 0600
boundaryapp-scoped
lifetime24 hours
forbiddenpasswords + copied secrets

Every change is a release you can inspect.

Artifacts are immutable, SDKs are pinned per deployment, migrations are append-only, and an explicit rollback restores the complete prior release.

2026.07.29-3active · sdk 0.2.1
2026.07.29-2superseded
2026.07.29-1available to restore

Your next app

Describe it. Deploy it. Review it.

Start with the prompt now. Your terminal agent can verify the CLI, register a private project when needed, build in isolated dev, and promote the exact passing revision before handing you the live app.

View CLI quickstart
✓ Prompt copied. Paste it into Codex, Claude Code, or another terminal-enabled agent.