Run OpenClaw on Curate-Me in 5 Minutes
OpenClaw is an incredible open-source AI agent — 215K GitHub stars and growing. But running it in production means dealing with security patching, cost control, and infrastructure ops. With the creator recently joining OpenAI and the project transitioning to a foundation, having a managed governance layer is more important than ever.
Curate-Me handles all of that. You keep your OpenClaw config. We add governance.
Step 1: Sign Up & Get Your API Key
- Go to dashboard.curate-me.ai/signup
- Create your account (no credit card needed)
- Copy your API key from the welcome page:
cm_sk_xxx
Step 2: Point OpenClaw at the Gateway
Change one environment variable. That’s it.
# Before (direct to provider — no governance):
OPENAI_BASE_URL=https://api.openai.com/v1
# After (through Curate-Me — full governance):
OPENAI_BASE_URL=https://api.curate-me.ai/v1/openaiAdd your Curate-Me API key as a header:
X-CM-API-Key=cm_sk_xxxYour existing OpenClaw config stays the same. Your agents don’t even know they’re being governed.
Step 3: Verify It Works
Python
from curate_me import CurateMe
client = CurateMe(api_key="cm_sk_xxx")
health = client.gateway.health()
print(health) # {"status": "ok", "gateway": "active"}TypeScript
import { CurateMe } from '@curate-me/sdk'
const client = new CurateMe({ apiKey: 'cm_sk_xxx' })
const health = await client.gateway.health()
console.log(health) // { status: "ok", gateway: "active" }curl
curl https://api.curate-me.ai/v1/health \
-H "X-CM-API-Key: cm_sk_xxx"Step 4: Configure Governance (Optional)
From the dashboard, you can enable:
- Cost controls: Set daily budgets per org ($25/day default)
- PII scanning: Block API keys and secrets before they hit providers
- Model allowlists: Restrict which models your agents can use
- HITL approvals: Require human approval for high-cost operations
- Rate limiting: Throttle requests per API key
All configurable from dashboard.curate-me.ai → Policies.
What You Get
| Feature | Self-Hosted | With Curate-Me |
|---|---|---|
| Cost tracking | None | Real-time per-model |
| PII scanning | None | Automatic |
| Budget caps | None | Per-org daily limits |
| HITL approvals | None | Approval queues |
| Security audit | None | Compliance scoring |
| Dashboard | None | 41-page ops console |