Control Panel
Guided deploys & one-click rollback
Collect artifacts from a local build, S3 or Jenkins, lay out service folders, install and start. Every deploy is recorded with source, build, environment and result.
New Git pull requests · load-test verdicts · AI assistant
One console for every
One console for every
environment you ship to
Control Panel deploys your builds, restarts your services, tails your logs, watches your uptime, load-tests your APIs and queries your databases — across Windows, Docker, Linux and Kubernetes. Run it as a hosted workspace, or install one binary on your own machine and stay entirely offline.
Free tier, no card Single binary, no agents Air-gapped installs supported
4
environment types: Windows, Docker, Linux, Kubernetes
16
database engines in the built-in client
1
binary — UI embedded, nothing to install alongside
0
outbound calls needed: licences verify offline
The tab problem
Operations is spread across six tools that don't know about each other
A deploy in Jenkins. Logs over SSH. Uptime in one SaaS, alerts in another. A JMeter install nobody maintains. DBeaver for the database, Postman for the API, and a wiki page explaining which PowerShell script to run on the Windows box.
Control Panel is one console over all of it — with the audit trail, the roles and the history that a stack of separate tools can never give you.
Before
Six tabs, three credentials, one person who knows the runbook. No record of who restarted what.
After
One URL. Pick the environment, see every service, deploy build 2481, watch it go healthy, and read the audit line that says you did it.
What's inside
An operations console, not a dashboard
Everything a DevOps team runs day to day, in the product today — one binary, no plugins to buy.
Uptime monitoring that alerts
HTTP/TCP checks with persisted history, 24h/7d uptime, latency sparklines and an incident timeline — alerting Slack, Teams, PagerDuty, webhooks or email on down and recovery. A check that never reached the service is held apart from one that failed, and a whole outage arrives as one message.
Multi-pane live logs
Follow any number of services side by side in a split grid, with search and highlight across every pane at once and per-pane match counts.
Load testing with a pass mark
A built-in HTTP engine with no JVM: stages, arrival rates and weighted scenarios, pass/fail criteria a
pipeline can gate on, a live dashboard, baselines and trends. JMeter .jmx still runs if you
have one.
16 databases, one client
PostgreSQL, MySQL, SQL Server, Oracle, SQLite, ClickHouse, Mongo, Cassandra, Redis, Elasticsearch, DynamoDB, Firestore, InfluxDB, Neo4j, Couchbase and SurrealDB — query, edit rows, import CSV and copy whole schemas between them.
A Postman-shaped API client
Auth, every body mode, {{variables}}, pre-request and post-response scripts, saved
collections, OpenAPI/Postman import — and requests issued from inside the target environment.
File explorer & web terminal
Browse, edit inline, upload, rename, chmod over a pooled SSH connection — plus an admin-only audited shell to a host or into a container, in split panes that each say which environment they are in.
DORA metrics for managers
Deploy frequency, change-failure rate, MTTR from the incident timeline, and 30-day SLA per target — read from what actually happened, not a spreadsheet.
Roles, SSO and an audit log
Four roles enforced server-side, narrowed per environment where you need it, OIDC single sign-on, and every login, command, file write and denied request appended as JSON lines with secrets redacted.
Git across every repository
Compare branches, then review, approve and merge pull requests on GitHub, GitLab and Bitbucket from one dashboard — and open the same change in many repositories at once.
Utilities that never upload
Encoders, a JWT decoder, image, icon, PDF (with Lock and Unlock), document and audio tools, and a multi-track Remix — all running in the browser, so nothing leaves the machine.
An assistant that looks it up
Ask about your environments and it answers from the panel's own endpoints, with your permissions and every lookup listed — off until an admin turns it on.
Deploy
From artifact to running service, with a way back
Pick where the build comes from, pick the environment, pick the services. The panel does the folder layout, the install and the restart — then records what it did so the next person doesn't have to ask.
- Sources — local build, S3 bucket or Jenkins job.
- Currently deployed build shown in the services header, so "what's on QA?" stops being a question.
- One-click rollback to the previous successful build.
- CI webhook — Jenkins or GitHub Actions POSTs an HMAC-signed body and the deploy lands in history.
deploy · STAGING
Deployment historyrollback ready
| When | Build | Source | By | Result |
|---|---|---|---|---|
| 12:04 | 2481 | jenkins #914 | admin | success |
| 09:38 | 2480 | s3://builds | ci-webhook | success |
| Tue | 2477 | local build | operator | rolled back |
| Mon | 2476 | jenkins #908 | admin | success |
Deploys / wk
17
Change failure
5.9%
MTTR
22m
Monitor
Know it broke before your customer tells you
Point a check at an HTTP endpoint or a TCP port and the panel keeps its history — uptime percentages, latency over time, and an incident timeline you can hand to a post-mortem.
- Alerts on down and recovery to Slack, Microsoft Teams, PagerDuty, a generic webhook, or email.
- Email templates — concise, detailed, incident report or HTML summary; start a custom one from any of them.
- Prometheus
/metricsso an existing Grafana stack sees the same numbers. - Fleet view — every target at a glance beside the live audit feed.
monitoring · fleet
Health checks 23 up · 1 degraded
| Target | 24h | 7d | Latency | |
|---|---|---|---|---|
| api.prod | 100% | 99.98% | 118ms | |
| checkout.prod | 99.9% | 99.91% | 204ms | |
| billing.prod | 97.2% | 99.40% | 1.9s |
incident timeline
09:14:02 DOWN billing.prod — 502 from upstream
09:14:03 ALERT slack #ops-alerts · pagerduty P-4192
09:36:41 UP billing.prod — recovered after 22m 39s
Operate
Logs, files and a shell — without opening a terminal
Tail four services at once and search across all of them. Edit a config file in place. Drop into an
audited shell on the host, or straight into a container with docker exec.
- Split-pane log grid with cross-pane search, highlight and match counts.
- SFTP-style explorer — inline edit, upload, download, rename, properties,
chmod. - Pooled SSH in the Go backend — no handshake per click, so browsing feels local.
- Named server-side actions — the browser never sends a shell string.
logs · 4 panes · "timeout"
gateway 2 hits
12:04 INFO routed /v2/pay → checkout
12:04 WARN upstream timeout (1 of 3)
12:05 INFO retry succeeded in 240ms
checkout-api 1 hit
12:04 INFO POST /charge 201 84ms
12:04 ERROR db timeout after 5s
12:05 OK pool recovered
billing-worker
12:05 INFO restarting (operator)
12:05 INFO jvm args from launch_args
12:05 OK started in 3.1s
search-svc
12:04 INFO index segment merged
12:05 INFO q="pric*" 12ms 41 hits
12:05 INFO heap 512M/1G
Environments
Agentless: it meets your infrastructure where it is
Agentless server management: no agent is installed on a managed host — the panel connects out over SSH, WinRM or kubectl. Define environments once in config or in the UI; everything else in the panel follows the one you have selected.
Windows
Local machine or remote over WinRM, with RDP for interactive sessions. Scheduled-task services driven by the panel's own Go code — no helper scripts to keep on disk.
Docker & Linux
Remote hosts over SSH — direct, or through a bastion / jump host. Container operations run as named server-side actions.
Kubernetes
Driven with kubectl from a jump host, a master node or the panel itself. Services are
Deployments; stop and start scale to zero and back to the previous replica count.
Desktop extras
On the machine the panel runs on: disk cleanup by category, photo, video and contact import from an iPhone or Android — audited, sync-style — and desktop recording to MP4.
Two ways to run it
Hosted when you want it easy. Yours when it has to be.
The same console either way. The difference is who holds the machine.
SaaS
Hosted workspace
Sign up, get a workspace, add your first environment in minutes. We run the server, keep it patched and back it up. Your workspace's configuration is isolated in its own tenant.
- Nothing to install or upgrade
- Per-user billing, cancel any time
- SSO, RBAC and audit export on Business
- Connects out to your hosts over SSH or WinRM
Self-hosted
Offline install, one binary
A per-user .msi, a .dmg, a container image, a Helm chart or a systemd unit.
The UI is embedded in the executable — copy one file anywhere and run it.
- Your data never leaves your network
- Licence verified offline with a pinned Ed25519 key
- Air-gapped activation by machine ID
- Host shell, filesystem and disk cleanup available
Host access is a self-hosted feature, on purpose. In a hosted workspace the shell, the host
filesystem and disk cleanup are not registered at all — a workspace admin is a stranger to our machine.
Self-hosted, the admin already owns the box, so those tools are exactly what the product is for.
Pricing
Start free. Pay when it's doing your job for you.
Per-user for the hosted workspace, per-install for a self-hosted licence. No per-host metering, no charge for the environments you monitor.
save 20%
Free
One person, one environment, everything working end to end.
$0forever
No card required
Create a workspace
- 1 user · 1 environment · 5 services
- 3 uptime checks, 7-day history
- Logs, files, API client, DB client
- Community support
Most teams
Team
A squad that deploys to dev, QA and prod and wants alerts that arrive.
$24/user / month
Billed monthly
Start 14-day trial
- Everything in Free, plus
- Unlimited environments & services
- Unlimited checks · 90-day history
- Slack, Teams, PagerDuty, webhook & email alerts
- Deploy history & one-click rollback
- Load testing & HTML reports
- Scoped API tokens, CI webhook,
cpctl
Business
Shared prod, more than one team, and a security review to pass.
$49/user / month
Billed monthly
Start 14-day trial
- Everything in Team, plus
- SSO via OIDC (Okta, Entra ID, Google, Keycloak)
- Per-environment role scoping
- Audit export to SIEM · 1-year retention
- DORA reporting & 30-day SLA per target
- Secrets-manager refs (Vault, env, file)
- Priority support, 1 business day
Enterprise
Procurement, an SLA, and infrastructure that has its own rules.
Talk
Annual contract · invoice
Contact sales
- Everything in Business, plus
- Self-hosted or single-tenant deployment
- Air-gapped & machine-bound licences
- Security review, SBOM, signed artifacts
- Uptime SLA & named support engineer
- Onboarding and migration help
Self-hosting instead? Licences start at $149 per install per year — see self-hosted pricing
Trust
Privileged tooling, treated as such
This console executes commands against your hosts. That is the product — so the controls around it are the first thing we build, not the last.
- Server-side auth and four roles — viewer, operator, deployer, admin — enforced in Go on every route, not in the UI.
- Audit log of every login, command, file write, config change and denied request, with credential-looking values redacted.
- No shell strings from the browser — the client names an action; the server builds the command.
- SSH host-key pinning, CSRF/origin checks, a filesystem sandbox, and secrets resolved at use-time from Vault or files.
- Offline licence verification — a pinned public key, so your install keeps working whether or not anything of ours is reachable.
Questions
Before you sign up
Do I need to install an agent on every host?
No. The panel connects out over SSH (with a pooled, persistent connection) or WinRM,
and drives Kubernetes with kubectl. There is nothing to deploy on the machines you manage.
Can I move between hosted and self-hosted later?
Yes — it is the same software. Settings can be exported and cloned into another install, so a workspace can become a self-hosted panel or the other way round. Host-level features (shell, filesystem, disk cleanup) only light up in a self-hosted install.
What happens to a self-hosted install if my licence expires?
It stops serving the console and shows the activation screen; nothing is deleted, and
your config/ and history stay on disk. Renew and it picks up where it left off. Verification is
offline, so a licence that is still valid keeps working even with no route to the internet — forever.
Does it work on an air-gapped network?
Yes. The activation screen shows a machine ID that carries no hostname or MAC address; send it to us and we issue a token already bound to that machine. It then verifies with no network at all.
Is Windows required?
Only for managing Windows targets, which the panel drives over WinRM. Docker, Linux and Kubernetes environments are Linux-native and run happily from the container image or a systemd unit.
How is this different from Grafana, Portainer or Rundeck?
Each of those does one part well. Control Panel is the operator's whole loop in one place — deploy, restart, tail, monitor, alert, load-test, query the database and record who did it — over mixed Windows/Docker/Kubernetes estates, where most tools pick a side.
Can I try it before paying anything?
Create a free hosted workspace with no card, or download the installer and press Start my 30 free days on its activation screen — the self-hosted trial issues itself, no sales call. Both run the full console; the free workspace is limited by size, not by feature.
Still deciding? Ask us anything — a real engineer answers.
Put your whole estate behind one URL
A single pane of glass for Windows, Docker, Linux and Kubernetes. Free to start, 14-day trial on the paid plans, and a self-hosted licence when your network says the data stays home.
$
docker compose build && docker compose up -d