Skip to Content
BlogFebruary 2026: Platform Launch

February 2026: Platform Launch

Published February 27, 2026

We are excited to announce the public launch of the Curate-Me platform — the governance layer for AI agents. Here is everything that shipped this month.


AI Gateway Proxy

The core product: a reverse proxy that sits between your agents and their LLM providers.

  • Provider support: 51 providers across 7 tiers including OpenAI, Anthropic, Google (Gemini), DeepSeek, Groq, Mistral, xAI, and more
  • Zero code changes: Swap one environment variable (OPENAI_BASE_URL) and you are done
  • SSE passthrough: Full streaming support with async response passthrough
  • Upstream resilience: Automatic retry with exponential backoff on provider failures

5-Step Governance Chain

Every request passes through a policy pipeline that short-circuits on first denial:

  1. Rate Limiting — Per-org, per-key request throttling
  2. Cost Estimation — Pre-flight cost check against per-request and daily budgets
  3. PII Scanning — Regex scan for API keys, passwords, and PII in request content
  4. Model Allowlists — Enforce which models each org or key can use
  5. HITL Gate — Approval queues for high-cost or sensitive operations

Real-Time Cost Tracking

  • Per-request cost recording (Redis accumulator + MongoDB audit log)
  • Daily budget enforcement with configurable limits per organization
  • Cost breakdown by model, agent, and provider
  • Cost velocity anomaly detection

Managed Runners

Secure OpenClaw sandbox containers with full lifecycle management:

  • 4-tier sandbox isolation: READ_ONLY, WRITE_PROJECT, WRITE_SYSTEM, FULL_ACCESS
  • Network phase separation: SETUP (network on) > EXECUTION (network off) > TEARDOWN (network on)
  • 97 backend modules: Lifecycle, security, networking, CI, compute, skills, and more
  • Desktop streaming: Real-time VNC streaming to watch agents work
  • Time-travel debugging: Replay and inspect any agent execution step-by-step
  • CI auto-fix: Classify failures, auto-diagnose, and create fix PRs
  • Git worktree isolation: Per-session branches, no cross-contamination

Dashboard (64+ Pages)

A full operations console covering every aspect of the platform:

  • Runner management: List, detail, actions, templates, schedules, triggers
  • Cost tracking: Real-time spend dashboards with historical trends
  • Security: Sandbox configuration, network phases, compliance scoring
  • Infrastructure: Headless CI, worktrees, MCP servers, hooks, subagents
  • Billing: Plan management, invoices, usage tracking

SDKs

Production-ready SDKs for every major runtime:

  • Python: pip install curate-me
  • TypeScript: npm install @curate-me/sdk
  • CLI: npm install -g @curate-me/cli
  • Embed widget: npm install @curate-me/embed

Multi-Tenancy

Full organization-based data isolation:

  • JWT-based authentication with org context
  • Tenant isolation middleware on all B2B routes
  • Per-org API keys with scoped permissions
  • Role-based access control (Owner, Admin, Member)

Pricing

Four tiers to match every stage of growth:

PlanPriceHighlights
Free$0/mo10K gateway requests, 1 runner, PII scanning
Starter$49/mo100K requests, 3 runners, model allowlists
Growth$199/mo500K requests, 5 runners, HITL approvals, priority support
Enterprise$499/moUnlimited requests and runners, SLA, dedicated support

What is Next

  • Webhook integrations for real-time event notifications
  • Custom governance policies for enterprise workflows
  • SSO/SAML for enterprise authentication
  • Your Machines — connect your own hardware and run agents locally with full governance

Stay up to date by following the changelog or emailing hello@curate-me.ai.