Production-Ready

Your agent deploys
its own infra.

One API call provisions compute, Postgres, Redis, object storage, and a public domain. Live URL back in ~60 seconds. Works with OpenClaw, Claude, GPT, and any AI agent.

Early access. No credit card. No spam.

saga orchestrator
$ brigent deploy --image nginx:alpine --db --cache
Validating request…
Estimating cost…
Provisioning Railway compute…
Creating Neon Postgres…
Spinning up Upstash Redis…
Configuring Cloudflare DNS…
Health check (attempt 1/20)…
→ Live at https://my-service-k7x.draftlabs.org
34 MCP Tools
7 Deploy Types
5 Cloud Providers
~60s To Live URL
How It Works

One call in. Live URL out.

Your agent sends a deploy request. Brigent's saga orchestrator provisions every resource in sequence — with automatic rollback if any step fails.

01

Deploy Request

Docker image or GitHub repo. Pick your deploy type: API, fullstack, frontend, worker, cron, static site, or database-only. Add resources: database, cache, storage.

02

Saga Orchestration

Railway compute, Neon Postgres, Upstash Redis, Cloudflare R2 & DNS — provisioned sequentially. Per-step timeouts. Reverse-order compensation on failure.

03

Health Check

Configurable health check path. 20 retries, exponential backoff, 5-minute timeout. When it responds, you get a public URL and a subdomain on draftlabs.org.

04

Full Lifecycle

Scale replicas 1-10, resize memory 256MB-32GB, stream logs via SSE, pause/resume to control costs, update env vars with auto-redeploy. All via API.

Your Agent OpenClaw · MCP · SDK
deploy
Brigent Saga · Billing · Security
provisions
Railway Compute
Neon Postgres
Upstash Redis
Cloudflare R2 + DNS

Automatic rollback if anything fails. Dead letter queue for edge cases. AES-256-GCM encrypted secrets injected into compute.

Use Cases

What your agent can build.

Real services, deployed and managed autonomously. No human clicking through dashboards.

deploy — api
> brigent deploy --type api --db --cache
Railway compute provisioned
Neon Postgres created
Upstash Redis ready
DNS configured
→ https://payments-api.draftlabs.org

Backend APIs & Microservices

REST or GraphQL endpoints with Postgres and Redis. Webhook receivers, data pipelines, auth services — live in 60 seconds.

Railway Neon Upstash DNS
> deploy --type fullstack --db

Full-Stack Web Apps

Dashboards, admin panels, internal tools with frontend, backend, database, and a custom subdomain.

Railway Neon DNS
> deploy --type worker --cache

Autonomous Agent Services

OpenClaw agents, trading bots, monitoring agents, scrapers — self-deploy and run 24/7 with log streaming and auto-scaling.

Railway Upstash OpenClaw
> deploy --type cron --db --storage

Scheduled Jobs & Pipelines

Nightly reports, data syncs, periodic processing. Ephemeral mode auto-terminates after 24h.

Railway Neon R2
> deploy --type static_site --storage

Static Sites & Docs

Marketing pages, documentation portals, status pages. Deployed to R2 with custom domain. Zero compute cost.

R2 DNS
> deploy --type database_only

Managed Databases

Standalone Postgres for shared state between agents. Connection strings, branch snapshots, health checks.

Neon
Capabilities

Built for agents,
not dashboards.

Every feature is API-first. Your agent calls endpoints. No clicking around in a console.

Deploy

7 Deploy Types

API, fullstack, frontend, worker, cron, static_site, database_only. One endpoint, one schema. Docker images or GitHub repos.

Reliability

Saga Orchestration

Sequential provisioning with per-step timeouts. If step 4 fails, steps 3→2→1 auto-compensate. Failed compensations go to a dead letter queue.

Billing

USDC Credits

Deposit via x402 on Base chain ($5/$10/$50/$100 tiers). Hourly metering. Zero balance auto-pauses all deployments — no surprise bills.

Observe

Real-Time Streams

SSE for deployment events and live container logs. Webhooks with HMAC-SHA256 signatures and retry with exponential backoff. Prometheus metrics endpoint.

Security

Zero-Trust by Default

AES-256-GCM secrets management. Pre-deploy security scanner (images, env vars, SSRF patterns). Scoped API keys with bcrypt hashing. Rate limiting per key.

Scale

Elastic Resources

1–10 replicas. 256MB–32GB memory. Live env var updates with auto-redeploy. Pause/resume to control costs. Ephemeral mode auto-terminates after 24h.

Give your agents
infrastructure superpowers.

Join the waitlist for early access to Brigent.

Production-ready. We'll email your API key when it's your turn.