Updated on by Susanna Fagerholm
Happy New Year! We kicked off 2026 with exciting updates across Cyclr.
Check out our latest product update on the biggest highlights from the last month. Including a new API endpoint and support for larger file imports in the GitHub Integration. We also launched a brand new product for AI interoperability: MCP PaaS. MCP PaaS (MCP Platform as a Service) is designed to help SaaS and software companies bring AI interoperability to their products rapidly and securely.
- Launch of MCP PaaS, a new product for AI interoperability.
- New Cyclr API endpoint to generate setup links for Account Connectors.
- Improved paging and GitHub template import limits.
- How to create your first MCP Server for your SaaS Application.
And much more!
Product Spotlight
- Introducing MCP PaaS for rapid AI enablement
- We’ve introduced a new product to Cyclr: MCP PaaS. With MCP PaaS, SaaS companies can create MCP Servers for agentic using their existing APIs, with controls for method exposure, data minimization, connection security, and transaction auditability.
- Cyclr API: New endpoint to create setup tokens for Account Connectors
- Users can now generate setup links for Account Connectors via a new API endpoint, making it easier to create set up links programmatically.
Improved Performance and Security
GitHub Integration: Support for larger template imports
- Users can now import Cyclr Templates that are over 1 MB in size from GitHub repositories back into a Cyclr Console using the GitHub Integration feature.
Cyclr Interface: Updated paging controls
- Paging controls in list views, for example Transactions, have been updated from a dropdown to a textbox, making it quicker to jump to a specific page when working with large datasets.
We have also built a selection of private connectors for our partners. If you’d like our team of experts to build a connector for you, get in touch.
You’ll find all the latest updates and highlights in this month’s Release Notes.
Video Spotlight: Create your first MCP Server for your SaaS Application
AI agents can only act on what you let them access, and that is exactly the problem MCP PaaS is built to solve. This month, our technical advocate Wilson walks through how to build an MCP Server with Cyclr’s MCP PaaS, turning a SaaS application’s APIs into controllable AI tools.
Along the way, Wilson shows how to choose which connector methods to expose, add security like OAuth 2, and create separate MCP Servers for different functions to support role-based access. He also covers multi-tenancy for managing customer accounts, plus transaction trail logs for auditing MCP requests end to end.
Blog Spotlight: Why Would You Want an MCP PaaS?
AI agents are quickly becoming a new class of user. For SaaS teams, the challenge is enabling AI interoperability without handing over unrestricted access to sensitive data and operations.
In this article, we break down what MCP PaaS makes possible, including controlled method exposure, tighter data boundaries, stronger authentication, and clearer audit trails.
Read the full article.
