We already built our own internal developer platform. Why switch to Bunnyshell?

Custom IDPs require continuous engineering investment to maintain. Bunnyshell replaces homegrown environment tooling with a managed platform that handles ephemeral environments, preview deployments, CI/CD integration, DORA metrics, and cost reporting — so your platform team focuses on higher-value work.

Does Bunnyshell work with Docker Compose?

Yes. Bunnyshell supports Docker Compose, Helm charts, Kubernetes manifests, and Terraform — you can use your existing IaC without rewriting it. Multi-component environments can mix formats within a single project.

Does Bunnyshell support Helm and Terraform?

Both are first-class citizens in Bunnyshell. You can deploy Helm charts directly, use Terraform modules for infrastructure provisioning, and mix both with Docker Compose and raw Kubernetes manifests within a single environment definition.

What cloud providers does Bunnyshell support?

Bunnyshell supports all major cloud providers — AWS, Google Cloud, and Azure — as well as on-premises Kubernetes clusters. You bring your own cloud infrastructure; Bunnyshell manages the environment lifecycle on top of it.

Is Bunnyshell secure?

Yes. Environments run in isolated Kubernetes namespaces on your own cloud infrastructure — your code and data never leave your cloud account. Enterprise plans include audit logs, advanced RBAC, and a dedicated support SLA.

Is there a free trial for Bunnyshell?

Yes. Bunnyshell offers a free trial so your team can evaluate the platform with your actual workloads. The Startup plan provides pay-as-you-go access at $0.007 per minute per active environment with no contracts or minimum commitments.

We already built our own solution. Why switch?

Most teams that built their own environment tooling spend 80% of their platform engineering time maintaining it. Bunnyshell replaces that maintenance burden so your platform team can focus on unique capabilities. Migration typically takes days, not months.

Does it work with Docker Compose?

Yes. Bunnyshell converts Docker Compose definitions to Kubernetes resources automatically — health checks, volumes, cron jobs, ingress, and HPA. No Kubernetes knowledge required from your developers.

What about Helm and Terraform?

First-class support for both. Define Helm charts and Terraform modules as components alongside Docker Compose services. Your existing infrastructure-as-code becomes your environment definition — mix and match in a single YAML file.

What clouds do you support?

AWS, Azure, GCP, DigitalOcean, Scaleway, DT Cloud, and any private Kubernetes cluster — including on-premise and air-gapped. Multi-cloud from day one.

Is there a free trial?

14-day full-feature trial, no credit card required. Includes a Bunnyshell-managed Kubernetes cluster so you can start immediately — no need to connect your own infrastructure on day one.

How does pricing work?

Starts at $49/month. Pay-as-you-go based on environment runtime. Stopped environments cost nothing. Auto-sleep and auto-destroy rules help control costs automatically.

SOC 2, ISO 27001, and ISO 9001 compliant. BYOC model — your code and data stay in your cloud. Granular RBAC with Policies, Resource Selectors, and Teams. Namespace isolation, secrets encryption, IP whitelisting.

$ bns environments list

Ship Software
10x Faster.

Modern CI/CD, cloud development, preview, and ephemeral environments — all in one platform built for modern engineering teams.

Start Free ↗Book a Demo

No credit card required

// See Bunnyshell in action in 3 minutes

environments.bunnyshell.com

Used by engineering teams at

$ cat how-it-works.md

How Bunnyshell works

Connect your repo

Link your GitHub, GitLab, or Bitbucket repository. Define your environment using Docker Compose, Helm, or Terraform — tools you already use.

GitHub

GitLab

Bitbucket

Docker

Helm

K8s

Open a pull request

Bunnyshell auto-provisions an isolated, full-stack environment for every PR. Share the preview URL with your team.

PR #142→env-142.preview.bunnyshell.dev• Running

PR #143→env-143.preview.bunnyshell.dev• Running

PR #144→env-144.preview.bunnyshell.dev• Running

Merge and move on

Tests pass, code is reviewed, PR merges. The environment auto-destroys. No cleanup. No surprise cloud bills.

✓End-to-end tests passed

✓Code review approved

✓Merged — environment destroyed

✦✦

See it in action.

Deploy your first environment in under 30 minutes. 14-day trial, no credit card required.

Start Free ↗Book a Demo

Unified Environment
Management.

Centralize control of all your environments with a single global view. Create, clone, start, stop, and deactivate environments through one intuitive interface. Full oversight across projects, teams, and clusters.

✓Faster onboarding — new developers start coding in minutes with standardized, preconfigured environments
✓Self-service autonomy — teams create, pause, or discard environments without waiting on DevOps
✓Consistent & reliable — every environment mirrors production, ensuring predictable results across the stack
✓Clone any environment — duplicate a running environment in one click — same config, same data, new namespace

Environments

AllPrimaryEphemeral

+ Create environment

Environment	State	Cluster	Last update
frontend-checkout	Deployed	Production-US	2 hours ago
api-gateway-v2	Deployed	Production-US	2 hours ago
dev-rewards	Deployed	Staging-EU	5 hours ago
dev-billing-ms	Running	Staging-EU	1 day ago
PR #312 — checkout	Running	Staging-EU	12 min ago
PR #310 — auth-flow	Running	Production-US	45 min ago
PR #308 — refactor	Sleeping	Staging-EU	3 days ago
dev-ai-assistant	Deployed	Staging-EU	12 days ago
staging-main	Deployed	Production-US	1 hour ago
PR #305 — hotfix	Destroyed	Staging-EU	5 days ago

10 results

Ephemeral
Environments.

Every pull request gets its own full-stack environment — your app, your database, your infrastructure. Isolated, production-like, destroyed on merge. No more "who broke staging?"

✓Auto-deploy on PR open — environment spins up in minutes, with a shareable preview URL
✓Auto-destroy on merge — no forgotten resources, no bill creep
✓Full-stack replication — not just frontend previews. App + database + infrastructure
✓Git ChatOps — /bns:deploy, /bns:stop, /bns:start from PR comments

See ephemeral environments →

bunnyshellbotcommented just now

✅ Preview Environment deployed on Bunnyshell

Component	Endpoints
argocd	https://argocd-nwmok2.bunnyenv.com/
argocd-ttyd	https://argocd-web-cli-nwmok2.bunnyenv.com/

See: Environment Details | Pipeline Logs

Available commands (reply to this comment):

🔴 /bns:stop to stop the environment
🚀 /bns:deploy to redeploy the environment
❌ /bns:delete to remove the environment

Remote
Development.

Code in a cloud replica of production from your local IDE. VS Code, JetBrains — real-time file sync, step-through debugging, zero local dependencies. Start coding in minutes, not hours.

✓Local IDE, cloud runtime — edit locally, changes sync instantly to your cloud environment
✓Remote debugging — step-through for Node.js, PHP, Java. Attach your debugger to cloud containers
✓Port forwarding — access any remote service locally, as if running on your machine
✓SSH access — drop into any container. Inspect logs, run commands, troubleshoot in real-time

See remote development →

Terminal

$ bns remote-development up --component YMtxx2Oelq

Local Sync Path /Users/dev/Projects/my-app

Remote Sync Path /usr/src/app

INFO: All files will be synchronized...

Pod is ready for Remote Development.

Bunnyshell Remote Development Environment

/usr/src/app/backend #

VS Code—/usr/src/app

● Remote: bunnyenv.com

Explorer

src/

app.ts

routes/

api.ts

auth.ts

models/

user.ts

config.ts

package.json

Dockerfile

app.tsapi.ts

1import express from 'express'

2import { router } from './routes/api'

3import { authMiddleware } from './routes/auth'

5const app = express()

6app.use(authMiddleware)

7app.use('/api', router)

9app.listen(8080, () => {

10 console.log('Server running on :8080')

11})

● Remote Dev ActivebackendTypeScript

Ln 7, Col 32UTF-8

Your entire stack,
in a single YAML file.

Define your app, databases, and infrastructure in bunnyshell.yaml. Version-controlled, reproducible, drift-free. Bunnyshell extends Infrastructure-as-Code to the entire environment — not just infra.

✓7 component types — Docker Compose, Helm, K8s Manifests, Terraform, Custom Docker Image, Static App, Generic
✓Dependency orchestration — dependsOn controls execution order; independent components deploy in parallel
✓Powerful variables — Twig-like interpolation with conditionals, filters, cross-component references
✓Variable Groups — share secrets and config across environments, override per-env

Docker

Kubernetes

Helm

Terraform

See Environment as Code →

bunnyshell.yaml

2 kind: Application

3 name: backend

4 gitRepo: 'https://github.com/my-org/api.git'

5 gitBranch: main

6 gitApplicationPath: /

7 dockerCompose:

8 build:

9 context: .

10 dockerfile: ./Dockerfile

11 ports:

12 - '80:8080'

13 deploy:

14 resources:

15 limits:

16 cpus: '2'

17 memory: '4096M'

18 hosts:

19 - hostname: 'api-{{env.base_domain}}'

20 path: /

21 servicePort: 80

Templates &
Service Catalog.

28+ pre-built templates or bring your own from Git. Any team member spins up a full stack from the catalog — no DevOps knowledge required. The easiest way to set up your environment.

✓Pre-built templates — Laravel, Next.js, Sylius, Magento, Django, PyTorch, and more
✓Custom templates from Git — enforce org standards, compliance requirements baked in
✓Template Sync — detect drift between environment and template with built-in diff editor
✓One-click deploy — developers, QA, and product managers self-serve without tickets

See templates →

Sylius 1.12 Development

Full Sylius ecommerce setup with Nginx, PHP-FPM, MySQL

PHPNginxMySQL

DeployView repo

Express + React + PostgreSQL + AWS S3 (Terraform)

Full-stack app with Terraform-managed S3 bucket

NodeReactPGTF

DeployView repo

Express + React + AWS RDS PostgreSQL (Terraform)

Cloud-native stack with managed RDS database

NodeReactRDS

DeployView repo

Express + React + PostgreSQL (Helm + K8s manifests)

Kubernetes-native with Helm chart deployment

NodeReactHelm

DeployView repo

PyTorch Hello World

ML model training environment with GPU support

PythonPyTorch

DeployView repo

Spring Boot + React + MySQL

Java backend with React frontend and MySQL database

JavaReactMySQL

DeployView repo

Workflow Pipelines.
Dependency-aware.

Every environment action — deploy, destroy, start, stop — runs as a workflow of stages, jobs, and steps. Bunnyshell reads your dependsOn graph, parallelizes independent components, and sequences the rest. Deploy and Start respect dependency order. Destroy and Stop run everything in parallel.

✓4 lifecycle workflows — Deploy, Destroy, Start, Stop — each auto-generated from your environment definition. No pipeline YAML to write
✓Dependency-aware staging — dependsOn controls execution order across stages. Independent components run in parallel within each stage
✓Real-time DAG visualization — stages as columns, jobs within stages, step-by-step output with commands, stdout/stderr, and exit codes
✓Smart deploy strategies — RollingUpdate (zero-downtime) or Recreate. Auto-detects PVC usage and forces Recreate when needed
✓Managed or self-hosted runners — control concurrency, timeouts, and retry behavior. Use Bunnyshell runners or bring your own

See workflow documentation → →

Deployment workflowSuccessful

Pipeline ID: 304409

✓Pre checks

✓Connect to cluster

→

✓Cloud access

✓Fetch Git latest SHA

✓Ensure env namespace

→

✓Build app

✓Build nginx

→

✓Deploy app

✓Create volumes

✓Deploy mysql

→

✓Finalize cleanup

→

✓Check env status

Job Pre checks

✅ 2026-03-13 11:01:22.572Z Run "check environment"

✅ 2026-03-13 11:01:23.104Z Validate cluster connectivity

✅ 2026-03-13 11:01:24.891Z Fetch Git SHA a3f7c2d

✅ 2026-03-13 11:01:26.337Z Build app — using cached layers (3/5)

✅ 2026-03-13 11:01:41.012Z Build nginx — image pushed to registry

✅ 2026-03-13 11:01:55.448Z Deploy app — RollingUpdate 1/1 ready

✅ 2026-03-13 11:01:58.221Z Deploy mysql — pod stabilized

✅ 2026-03-13 11:02:10.667Z Finalize cleanup — removed 0 orphaned resources

Application & K8s
Logs.

Three log types, one interface. Pipeline logs show every workflow step with auto-expanded errors. Application logs stream live from running containers. Kubernetes logs surface pod events, manifests, and cluster-level diagnostics — no kubectl required.

✓Pipeline logs — every workflow job with timestamps, stdout/stderr, exit codes. Failed steps auto-expand and highlight in red
✓K8s Logs — real-time container output streaming. Filter by pod and container. Search across log lines
✓K8s Events — pod scheduling, image pulls, crash loops, OOM kills, restarts — all surfaced without cluster access
✓K8s Manifests — inspect the full YAML manifest for any running resource. Debug configuration drift in real-time

See troubleshooting docs → →

Acme Corp > staging > rabbitmq

rabbitmqRunning

RestartShell

K8s LogsK8s EventsK8s Manifest

Pods:All podsContainer:All containers

Streaming...

2026-03-13T14:22:01.112Z accepting AMQP connection <0.1234.0> (10.0.2.15:43210 -> 10.0.1.5:5672)

2026-03-13T14:22:01.118Z connection <0.1234.0> (10.0.2.15:43210 -> 10.0.1.5:5672): user 'guest' authenticated

2026-03-13T14:22:01.124Z connection <0.1234.0> (10.0.2.15:43210 -> 10.0.1.5:5672) has a client-provided name: app.consumer

2026-03-13T14:22:05.201Z accepting AMQP connection <0.1235.0> (10.0.2.16:43211 -> 10.0.1.5:5672)

2026-03-13T14:22:05.208Z connection <0.1235.0>: user 'guest' authenticated and granted access to vhost '/'

2026-03-13T14:22:10.334Z Queue 'tasks.default' in vhost '/': consumer added, 3 consumers total

2026-03-13T14:22:15.402Z accepting AMQP connection <0.1236.0> (10.0.3.20:52100 -> 10.0.1.5:5672)

2026-03-13T14:22:15.410Z connection <0.1236.0>: user 'app_worker' authenticated

Built for Scale.
Not hello-world demos.

Run thousands of environments without crashing your cluster. Smart parallel deploys for complex apps with 100+ microservices. Bunnyshell deploys real-world applications with dependency control and full visibility.

✓Dependency chains — use dependsOn to control which services launch first — great for DBs, caches, legacy systems
✓Dynamic concurrency — set how many environments run in parallel — match cluster capacity in real-time
✓Built-in queuing & retries — handle quick-succession commits without pipeline collisions
✓Fully configurable — order, concurrency, resources, and visibility — all as code

See enterprise capabilities → →

Acme Corp > staging > Components

stagingStopped

DeployStartConfigurationURLs

⚠Warning: There is a new version of the template used by this environment. Review changes

Histio-configs

acme/platform-inframaina3c81f2Stopped

Histio-routing

acme/platform-inframaina3c81f2Stopped

Histio-ingress

acme/platform-inframaina3c81f2Stopped

Hmysql-8

acme/platform-inframainb7d2e4aStopped

Hkafka

acme/platform-inframainb7d2e4aStopped

istio-configs

Pipeline LogsConfigurationExported Variables

✓ Pipeline completed — all steps finished

Automate everything.
Full API & CI/CD control.

Use your stack, your pipelines, your rules. Trigger PR environments with GitHub Actions, CLI, or ChatOps — no lock-in, no friction. Common setup: GitHub PR → CI job → Call Bunnyshell API → Get live URL → Run Playwright tests.

✓Full API control — create, update, delete environments programmatically. Open-source Go SDK
✓GitHub Actions & GitLab CI — native integrations — trigger environments as part of your existing CI flow
✓Built-in queuing — handle quick-succession commits without collisions. Automatic retries on failure
✓Environment ready in seconds — even for complex apps — parallel deploys, queuing, and CI-built images supported

Explore the API docs → →

Ephemeral environments

Automate environment creation from pull requests.

Create ephemeral environments on Pull Request

Auto-deploy a new environment when a PR is opened

Destroy environment after Merge or close Pull Request

Auto-destroy when the PR is merged or closed

REST API & CLI

# Create environment via API

curl -X POST https://api.bunnyshell.com/v2/environments \

-H "Authorization: Bearer $BNS_TOKEN" \

-d '{"name": "pr-142", "template": "tmpl_abc"}'

# Deploy via CLI

$ bns environments deploy --id env_12345

# ChatOps from PR comments

/bns:deploy /bns:stop /bns:start /bns:delete

GitHub Actions

- name: Deploy Preview Environment

uses: bunnyshell/deploy-action@v2

with:

access-token: ${{ secrets.BNS_TOKEN }}

environment-id: ${{ steps.create.outputs.env_id }}

Your terminal. Your IDE.
Your way of working.

CLI

Create, clone, start, stop, deploy, port-forward — all from your terminal. Shell autocomplete, profiles, secrets management.

API

REST API & Go SDK

Full programmatic access. Provision environments, query status, trigger teardowns — integrate with any workflow.

MCP Server

AI-native environment management. Create, deploy, and manage environments from Claude Code or any MCP-compatible agent.

Git

Git ChatOps

/bns:deploy, /bns:stop, /bns:start, /bns:delete — manage environments from PR comments. Auto-posted URLs and status.

DNS

Custom Domains

Assign custom domains to any environment. Per-component URL customization, load balancer config, IP whitelisting.

CI/CD

Webhooks & CI/CD

Trigger environment actions from GitHub Actions, GitLab CI, or any CI system. Fully scriptable, event-driven.

✦✦

Deploy your first environment today.

14-day trial, no credit card. Free white-glove onboarding — we set it up with you. Most teams ship their first PR environment in under 30 minutes.

Start Free ↗See It Live (15 min)

Teams that stopped
fighting over staging.

“Manual tests on Argo CD PRs now take me seconds. When you review dozens of PRs a day, that's a game-changer.”

Michael CrenshawCore Team Member

“Automatically created BunnyShell environment based on PRs, speeding up the entire task implementation process.”

Jacob TobiaszCore Team Member

Laura MichadCTO

“From a release once in 2-4 weeks to multiple times per week, and nobody's stressed about it.”

Catalin SporisCEO

“I can count on an integrated platform for managing our infrastructure — capabilities we don't have in-house, at a fraction of the cost.”

Gustavo HaramuraTeam Lead & Solution Architect

“I can do with Bunnyshell and a DevOps person what my clients are doing with 4 DevOps people.”

Cost & Lifecycle
Management.

Environments that manage themselves. Auto-sleep, auto-destroy, auto-update — so you stop paying for idle resources and your team never has to remember cleanup. Detailed cost analysis with grouping, filtering, and comparative charts.

✓Auto-sleep schedules — stop at 8 PM, start at 8 AM. Custom per environment
✓Auto-destroy on merge — no forgotten resources, no surprise cloud bills
✓Per-environment costs — Kubecost integration shows what each environment costs per hour. Group by project, cluster, or team
✓Termination protection — lock critical environments against accidental deletion

See cost management docs → →

Environment availability rules

Auto-stop

hours of inactivity

Timeframes

Define when environments should be running. Outside these windows, environments will be automatically stopped.

Monday – Friday08:00 AM→08:00 PM

Saturday10:00 AM→04:00 PM

Sunday—→—

+ Add timeframe

Auto-destroy on merge

Destroy environment when pull request is merged or closed

Termination protection

Prevent accidental deletion of this environment

Environment Drift
Management.

Get notified when your environment configuration diverges from the template. Review changes through a built-in diff-editor, adjust settings, or acknowledge changes without modification. Keep environments aligned with the latest template versions.

✓Auto-detection — Bunnyshell notifies you when an environment drifts from its template
✓Built-in diff editor — review exactly what changed — line by line, component by component
✓One-click sync — pull template updates into your environment or acknowledge the divergence
✓Version tracking — see which template version each environment is running against

bunnyshell.yaml

CopyShortcuts

1 kind: Environment

2- name: pr-demo-1

3- type: ephemeral

5 components:

6 - kind: Helm

7 name: api

8- version: 1.2.0

9- replicas: 1

11 environmentVariablesGroups:

12 common:

13 ENV_DOMAIN: acme-corp.app

14- LOG_LEVEL: debug

15 CACHE_TTL: 300

1 kind: Environment

2+ name: staging

3+ type: primary

4+ urlHandle: staging

5 components:

6 - kind: Helm

7 name: api

8+ version: 1.3.0

9+ replicas: 2

11 environmentVariablesGroups:

12 common:

13 ENV_DOMAIN: acme-corp.app

14+ LOG_LEVEL: info

15 CACHE_TTL: 300

CopyChanges reviewed

Enterprise-grade security.
Your data stays in your cloud.

Bunnyshell connects to YOUR Kubernetes cluster. Your code and data never leave your infrastructure. SOC 2, ISO 27001, and ISO 9001 compliant.

RBAC

Three-layer permission model: Policies, Resource Selectors, and Teams. Granular enough for enterprise.

SSO

Enterprise single sign-on. Integrate with your existing identity provider.

Network Isolation

Dedicated Kubernetes namespace per environment. Pods accessible only within their namespace.

Secrets Management

Environment variables and secrets handled securely. Never hard-coded, always encrypted.

IP Whitelisting

Restrict access to specific IP ranges. Control who can reach your environments.

Audit Logs

Full audit trail for every environment action. Who deployed what, when, and where.

✦✦

Enterprise-ready from day one.

SOC 2 compliant. BYOC model. RBAC, SSO, audit logs — all built in. Book a demo to see how Bunnyshell fits your security requirements.

Start Free ↗Book a Demo

DORA Metrics,
Built In.

Prove your team is shipping faster. Track Cycle Time and Deployment Frequency across your repositories — without buying another tool. PR-based or tag-based release detection, monthly breakdowns, and per-repo analytics.

✓Average Cycle Time — measure how long from first commit to production deploy
✓Deployment Frequency — track how often your team ships — daily, weekly, monthly trends
✓Cycle Time Breakdown — see where time is spent — coding, pickup, review, deploy
✓Per-repo analytics — compare performance across repositories and teams

Bunnyshell DORA metrics — Cycle Time, Deploy Frequency, monthly charts and per-repo analytics

Full Configurability.
Per Environment.

Every environment has granular settings you can tune independently. From builder infrastructure to deployment runners, from workflow parallelism to custom URL handles — configure exactly what you need without touching other environments.

✓Builder Infrastructure — choose managed or self-hosted builders per environment
✓Workflow Settings — parallel jobs, status check timeouts, image pull policies
✓Environment Variables — inherited from project, overridable per environment
✓URL Handle — custom domains and URL handles per environment

Acme Corp > staging > Settings

General Settings

Build Settings

Deployment Runner Settings

Workflow Settings

URL Handle

Environment Variables Groups

Environment Variables

Availability Rules

Ephemeral Environments

Build Settings

Builder Infrastructure

Project Default

Managed Builder

Self-Hosted Builder

Container registry

Project Default (Bunnyshell Registry)

Build Engine

Project Default (BuildKit)

12 Component Types.
One unified platform.

Docker

Docker Compose

Automatic conversion to Kubernetes resources. Health checks, volumes, cron jobs, ingress, and HPA — all from your docker-compose.yaml.

App

Application

Full application lifecycle management. Build, deploy, start, stop, and destroy with dependency orchestration.

Svc

Service

Long-running services with custom deploy, start, stop, and destroy scripts. Full control over lifecycle.

Static

Static App

Build and serve static sites. Node.js build pipeline with customizable output directory and build arguments.

Image

Docker Image

Deploy pre-built container images directly. Skip the build step when you already have CI-built images.

Helm

Helm Chart

Deploy Helm charts from any repository. Full Helm lifecycle with custom values and version pinning.

K8s

K8s Manifest

Apply raw Kubernetes manifests with kubectl. Kustomize support, custom scaling, and full RBAC control.

Terraform

Provision cloud infrastructure alongside your app. S3 buckets, RDS databases, CDNs — all as code.

Gen

Generic

Run any script on any runner image. Cloud Run, Lambda, custom CLIs — bring your own deployment logic.

Neon

Neon (Postgres)

Native Neon database branching. Each environment gets its own isolated Postgres branch — instant, zero-copy.

PlanetScale (MySQL)

Native PlanetScale database branching. Schema-aware MySQL branches per environment with zero data copying.

Xata

Xata (Data)

Native Xata database branching. Serverless data platform with per-environment isolated branches.

Kubernetes-Native
Observability.

Real-time K8s logs, events, and manifests for every component. Stream container logs live, inspect pod events, and view raw manifests — all from the Bunnyshell UI. No kubectl required, no cluster access needed.

✓Real-time log streaming — stream container logs live with pod and container filtering
✓K8s Events — see pod scheduling, image pulls, crashes, and restarts in real-time
✓K8s Manifest viewer — inspect the actual running manifest for any component
✓Multi-pod support — aggregate logs from all pods or filter by specific pod and container