BluBees Beta is Live. Here's Everything That's New

When we launched BluBees on the Atlassian Marketplace, we knew the real work was just beginning. Building something genuinely useful means getting it in front of real teams, watching how they use it, and iterating on what actually matters.

That's what Beta is for. And today, we're sharing everything that's in it.

This isn't an incremental update. It's a meaningful step forward in what BluBees can do, built directly on feedback from our Beta cohort and enterprise design partners. More reach for your AI Agents, smarter workflow logic, deeper integrations, and a canvas that's genuinely easier to use. Here's everything that's new and why it matters.

MCP Support: Your AI Agents Just Got Unlimited Reach

This is the biggest capability addition in this release.

Model Context Protocol (MCP) is the open standard that defines how AI models communicate with external tools. Until now, a BluBees AI Agent could only connect to tools that had pre-built Connectors in the platform. With native MCP support, that limitation is gone.

Any application that exposes an MCP server is now instantly available as a tool skill for any BluBees Agent, without custom integration work, without code, without waiting for a Connector to be built. Configure multiple MCP servers simultaneously, and your Agents can reach across your entire toolstack with the same consistency and governance you'd expect from a pre-built integration.

For enterprise teams, MCP also means centralised governance; the integration between the AI Agent and the MCP server enforces a single, controlled access point for all tool interactions. That matters for security, compliance, and auditability.

MCP doesn't just extend BluBees. It makes the boundary of what BluBees can reach effectively limitless.

New AI Models: DeepSeek and Anthropic Claude

We've expanded our LLM ecosystem significantly. BluBees now supports DeepSeek and Anthropic Claude natively, alongside the existing OpenAI GPT-4o and Google Gemini integrations.

Why does this matter? Because different tasks benefit from different models. A workflow step that needs deep reasoning performs differently with Claude than with GPT-4o. A high-volume data processing task has a different cost profile with DeepSeek than with Gemini. Multi-LLM support has always been part of BluBees' architecture — this release delivers on that promise with four production-ready model options.

Choose your model per Agent, per Flow, per use case. The right intelligence for the right job.

Webhook Node: Connect Anything With an Endpoint

Not every tool in your stack has a pre-built Connector. The Webhook Node closes that gap.

Drop it anywhere in any Flow — beginning, middle, or end. Configure the destination URL, authentication headers, and connection parameters directly in the node. When the Flow reaches that step, it fires the webhook, captures the response, and passes the payload downstream to whatever comes next.

The Webhook Node also enables Flow-to-Flow invocation, where in one Flow can trigger another through a webhook, opening up modular workflow architectures where complex processes are broken into smaller, reusable Flows that call each other as needed.

If a system has a webhook endpoint, BluBees can now talk to it. That's most of the internet.

Visual Conditional Route Builder: Workflows That Handle Reality

Real-world data doesn't arrive clean. Fields are missing. Values are unexpected. Edge cases appear at the worst possible moment.

The Visual Conditional Route Builder lets you define exactly how your Flows should respond to that reality. Instead of a simple field mapping where Field A always goes to Field B, conditional routes evaluate data at runtime and make decisions, route this record differently based on its priority field, handle this missing value with a default, send this data to a different endpoint if the approval field is No.

The result is workflows that are genuinely resilient, not brittle automations that break the moment data deviates from the expected pattern, but intelligent Flows that handle unpredictability as a feature, not an exception.

Centralised Secured Gateway for TestRay

For teams using TestRay alongside BluBees, connection just got simpler. Generate your TestRay API token directly within BluBees and connect to TestRay on the same Atlassian instance, without external configuration or credential management outside the platform. One secure gateway, managed from within BluBees.

Enhanced UI/UX: A Canvas Built for Everyone

We've overhauled the workspace canvas with one goal: making BluBees self-describing for non-technical users. You shouldn't need a technical background to understand what's happening in a Flow or to manage it confidently.

Here's what changed:

• Live Log download: export the current log as a plain text file directly from the Live Log window, useful for debugging, auditing, and sharing with support

• Clarified status controls: Flows are now clearly labelled Draft (not running) or Active (running), no ambiguity about what's live

• Tile View dashboard: active Flows are displayed as visual tiles, making it easier to manage multiple Flows at a glance

• Right-click context menu: right-click any node on the canvas to access configuration, view full step details, and manage the node without navigating away from the Flow

These aren't cosmetic changes. They're the difference between a tool that technical users can operate and a platform that everyone on the team can own.

Stability and Performance

Alongside new features, this release includes targeted fixes and performance improvements:

• Fixed a canvas layout bug where workflow nodes would occasionally render off-screen in negative coordinates

• Optimised app icon storage, migrated from Base64 strings to clean URLs, significantly improving API response times

• Implemented a unified error handler that surfaces external API downtime directly within the canvas view, no more silent failures

What's Coming Next

We're moving toward General Availability. Our engineering team is actively building features that will automatically discover your enterprise tech stack and map out recommended configuration Flows across your business tools simultaneously, reducing the time from "I want to automate this" to "this is running" dramatically.

We're currently in Beta, working closely with select teams to test and refine the platform. If you'd like to be part of the Beta cohort, reach out, and if you're already with us, everything in this release is live now.

Found a bug or have a suggestion? Our support portal is open: blubees.atlassian.net/servicedesk/customer/portal/1

Interested in joining the BluBees Beta? Visit blubees.ai or install the app from the Atlassian Marketplace to get started.