Access, auth, and limits
There is nothing to authorize yet. The route that works today is a Loops API key against the REST API documented at loops.so/docs, combined with the CLI skills Loops publishes for coding agents. That gives an agent the same reach the planned server describes, minus the runtime discovery and the scoped-consent model MCP brings with it.
When the server does ship, the announced surface covers contacts, events, transactional email, campaigns and workflows. Worth watching: transactional email and campaigns are both write paths that reach real inboxes, so whether Loops splits read and write consent the way Customer.io does will matter more than the tool count.
Loops also publishes a REST API
Loops publishes an API, but per-tier gating wasn't extracted automatically.
Limits and gotchas
- Fallback: the REST API works for any flow the MCP server doesn't cover.
- OpenAPI spec is published — easy to validate which endpoints the MCP server actually exposes vs the full API surface.
- No MCP available — until Loops ships one, use the REST API directly.
Agent-readiness verdict
Announced, not shipped. Treat any listing claiming a live Loops MCP server as stale, and build against the API with the CLI skills in the meantime.
Strong agent-fit for SaaS workflows. Loops has a community MCP server, documented public API with OpenAPI spec, and SDKs across Node, Go, Ruby, and PHP, plus native Claude integration. The main limitation is lack of webhook support and structured JSON outputs, which narrows real-time automation possibilities.
Scored by Joonas (TMB).
Loops MCP & API FAQ
Does Loops have a public API?+
Does Loops publish an OpenAPI / Swagger spec?+
Does Loops plan to add an MCP server?+
Sources
- Loops official site: https://loops.so
- API docs: https://loops.so/docs/api-reference/intro
Data verified by Joonas on the dates shown. MCP server status auto-rechecked weekly.
10+ years in digital marketing. I review marketing software for AI-stack fit: real pricing, MCP and API support, and how cleanly each tool drops into an AI agent workflow, cross-checked against verified data and real user feedback.

