Developer tools decision
Claws.host
A technically capable individual can reproduce the core offering (single dedicated agent instance, onboarding, terminal, and basic monitoring) in about a week; the vendor mainly packages managed provisioning, security hardening, and support which are reproducible but convenient to pay for.
Visit website↗$29/mo
$348/yr
Read off the official pricing page.
$50one-off23 h to build
$0/mo3 h/mo upkeep
On cash alone, building overtakes the subscription at 1 seat.
No open-source build does this yet
Nothing published replaces this one, so a replacement starts from an empty file. Here is what it would have to cover.
What a replacement has to do
- Provision a single dedicated VM, install and secure an agent runtime (OpenClaw / Hermes), configure the user's model API key, expose a web terminal and basic dashboard, and run basic monitoring/log streaming.
What it still won’t have
- Fully managed provisioning and automated instance creation on checkout
- Continuous OS-level patching and security-hardening provided by vendor
- Hosted billing, priority support, and vendor-run monitoring
- Turnkey onboarding wizard and one-click provisioning in-browser
What remains hard
- Execution quality
Your environment is provisioned on checkout, security-hardened at boot (UFW, fail2ban, SSH key-only, unattended-upgrades), and kept patched — you never touch infrastructure.
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 1 seat.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a single-tenant, self-hosted OpenClaw server stack using Ubuntu 22.04, Docker Compose, and a small Node.js + React dashboard. Core features in scope: (1) Terraform or cloud-init script to provision a VM and attach a public IP; (2) Docker Compose manifest to run OpenClaw (or Hermes) and a reverse-proxy; (3) security-hardening scripts (UFW, fail2ban, SSH key-only, unattended-upgrades) run at first-boot; (4) onboarding UI to accept and securely store the user's model API key and start the agent service; (5) browser-accessible web terminal (gotty/wetty) and a dashboard page showing CPU/memory/disk and start/stop controls; (6) basic log streaming and a health-check endpoint. Out of scope: multi-tenant billing, paid support, SLA-backed monitoring, and a marketplace for agents. Require input validation, error handling, unit tests for onboarding/server logic, and a README with deployment and rollback steps.
How we checked
How the score was reached
- Build verdict base78
- 2 cited sources+1
- Price verified on pricing page+3
- 3/3 assessment runs agreed+4
- Evidence score86
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time - so the same evidence always produces the same number.
How scoring works →Cited sources · 2
Every page the run actually retrieved.
- official productClaws.host | Run OpenClaw in 5 minutes, not 5 hours
- official pricingClaws.host Pricing
Integrity checks
What held up, and what did not.
