AI work orchestration with governance built in
Curate-Me helps teams assign, schedule, and govern recurring AI work across agents, templates, and direct LLM gateway traffic. Start with the gateway, add Autopilot templates, and expand into managed runners when your workspace is enabled.
Current availability, June 2026. The gateway is generally
available for governed LLM traffic. Autopilot templates
and report workflows are available for enabled workspaces. Managed runners,
BYOVM, agent fleets, and desktop streaming remain private beta behind
FF_GATEWAY_MANAGED_RUNNER_BETA / RUNNER_BYOVM allowlists.
Request access if you want the full team workflow.
What the platform coordinates
Pick the work you need done. Curate-Me gives it a governed execution path.
| Surface | What it handles |
|---|---|
| Gateway | Governed LLM proxying, cost caps, PII scanning, rate limits, approvals, audit |
| Autopilot Templates | Repeatable work such as code review, security audits, finance reports, support triage, and research briefs |
| Teammates & Roles | Named AI teammates with budgets, tool access, and workplace channels |
| Managed Runners | Private-beta OpenClaw containers, BYOVM, files, sessions, and desktop streaming |
See What Curate-Me Does for the product overview.
How it works
- Start with the gateway — swap your provider base URL and add a Curate-Me key
- Choose the work pattern — direct API traffic, an Autopilot template, or a teammate role
- Connect your tools — link Slack, Teams, GitHub, email, or provider secrets
- Review the output — every request carries cost, approval, and audit context
Already using an LLM provider directly? The gateway is a drop-in proxy. Swap your base URL, add your gateway key, and every existing call gets cost tracking and policy enforcement automatically.
What teams accomplish
- Code and DevOps — review PRs, triage CI, draft fixes, and produce report-backed recommendations.
- Security and compliance — scan repositories, assemble evidence packs, and route risky output for sign-off.
- Finance and operations — investigate spend, summarize variance, and generate recurring reports.
- Research and growth — build market briefs, sales research, SEO outlines, and news digests.
Every action is governed
Every LLM request runs through a 15-stage governance chain:
| Cost tracking | Real-time spend per teammate, per model, with daily budget caps |
| PII scanning | 33 patterns + optional NER — block, redact, or log |
| Rate limiting | Per-key and per-org RPM limits with IETF headers |
| Approval gates | High-cost or sensitive requests pause for human sign-off |
| Model allowlists | Restrict which models your team can use |
| Security scanning | Prompt injection, jailbreak, and exfiltration detection |
| Full audit trail | Every action logged with trace IDs for compliance |