Changelog

New updates and improvements to Runbear.

🚀 Better Smart Replies, Web Search Update, More Reliable Knowledge Base

Better Smart Reply

Smart Reply just got smarter! 🎉 Runbear AI Agents now proactively offer relevant information when they find it—no extra searching needed. This way, your teammates get answers faster, without interruptions.

Other changes

  • 🌐 OpenAI Web Search Update: We’ve upgraded from Bing Search to Serper API, ensuring continued support after Bing’s retirement (Aug 11, 2025). Serper provides Google-powered, AI-optimized results for more accurate and reliable answers.
  • 📂 Knowledge Base Improvements: Easier, faster, and more reliable

GPT-5, OpenAI MCP Support, Smarter Task Analysis, and More 🎉

🧠 GPT-5 Is Here Now at Runbear

Runbear now supports GPT-5 — OpenAI’s most advanced model yet, combining faster, smarter responses with built-in thinking for complex reasoning. Power your Runbear Agents with GPT-5! 👉 Learn more about GPT-5 in our deep dive and see its performance in our benchmark post.

🚀 Runbear OpenAI Agent Now Available via MCP

Your Runbear OpenAI Agent now supports MCP (Model Context Protocol), enabling seamless connections to external tools and data sources. — making your OpenAI-powered Agent smarter.

🔍 Smarter Task Analysis — From Slack Conversations to Ready-to-Use Agents

Task Analysis has been upgraded to be smarter and more actionable.

It now goes beyond detecting repetitive work — analyzing the context from your Slack conversations to understand your tasks and offer ready-to-use Agents.

Other changes

  • Trigger Agents in Zendesk with Specified Keywords: Call an Agent by typing a preset trigger word in an internal comment
  • Smart Reply Prompt: Include sample questions as prompts so Smart Reply is triggered more accurately when users want it
  • Custom MCP with Bearer Authentication: Now you can securely connect to MCP servers that use Bearer token authentication, enabling safe integration with token-based MCP servers.

Runbear Highlights: Zendesk Internal Answers & Anthropic Tool Logs and more!

🤖 Ask questions, get answers — all within Zendesk

You can now ask your own AI agent directly inside Zendesk and get instant answers without switching tools. From pricing policies to onboarding steps, resolve questions fast and send helpful replies to customers right away. ⚙️ Make sure to enable Trigger by internal comment in the Trigger tab to activate this feature.

✨Other changes

  • 📈 Per-Assistant Analytics Now Available: You can now view analytics by assistant in the Analytics tab on the agent edit page.

New in Runbear: SharePoint Sync, Scheduled Jobs, and Task Analysis (Beta)

🗂️ SharePoint Integration Is Live

You can now sync documents from SharePoint — like internal docs and team knowledge — so your agent stays aligned with what your team knows.

 🕒 Schedule Recurring Tasks in Slack

One of our most-requested features is now live: Scheduled Jobs.

You can now schedule recurring tasks directly from Slack — like check-ins, reports, or summaries — and let your agent handle them on time, every time.

Available now for Slack with Anthropic-powered agents.

Task Analysis Is Now in Beta

We’re launching Task Analysis — a smarter way to find work your agent can take off your plate.

It reviews Slack channels to uncover repetitive tasks and suggests agents to automate them — even the ones you didn’t know could be automated.

You’ll find it under Home > Task Analysis.

Note: Available on the Team plan and above. Slack only for now.

✨ Other Changes

Smarter agent creation: Just describe your agent’s goal — Runbear will generate a goal-optimized system prompt for you. You can further refine it with the ✨ magic wand button.

Improved tool execution flow: We’ve enhanced how agents call integrated tools, making task execution faster and more reliable across connected services.

Website Sync, Claude Web Search, and More!

🌐 Website Sync Has LandedOne of the most requested features is finally here!

You can now sync your website — including docs, blogs, and FAQs — directly to your agent’s knowledge base. It just works, and your agent instantly becomes more helpful.

We’re launching this in beta, and we’re excited to see what you’ll build with it.

🌎 Claude Agents can now Search the Web

Claude-powered agents can now search the web in real time — no setup needed.

That means they can pull the latest information straight from the internet whenever it’s relevant, and respond with even more accurate, up-to-date answers. This is now on by default.

Improved Reasoning, Tool Access Control, and Smarter MCP Integrations

Empower Connected MCP Tools

We’ve made the MCP tools smarter! They can now better understand which tool and settings you need—whether you’re adding blocks to a Notion page or creating and assigning Jira tickets.

Smarter Planning with Extended Thinking

Claude-based Agents are now better at handling complex tasks by thinking through the steps before taking action. This makes responses more accurate and reliable—especially for workflows that involve multiple decisions or tools.

Improved Reasoning Visibility

We’ve updated the response format for Claude-based Agents to make the reasoning process more transparent and easier to follow. Now, step-by-step thoughts remain visible throughout the response generation, giving users a clearer view of the full context.

Granular Tool Control for MCP Integrations

You can now selectively enable or disable tools within each MCP server. With this update, you can confidently empower AI agents to use only the tools you select, giving you more flexibility and control!

Private Beta: AI Continuous Learning with User Feedback

We’re excited to introduce a new continuous learning feature in private beta! The AI will automatically incorporate your feedback and adjust its system instructions in real time.

We’re rolling out this capability gradually to a select group of users. If you’d like early access to try it out, just reply to this email—we’d love to include you!

Smarter Replies in Teams DMs, and More!

More Natural Conversations in Microsoft Teams DMs

Agents can now understand and respond based on previous messages in 1:1 chats on Microsoft Teams. Previously, context was only maintained in threaded channel messages — DMs were limited to one-off replies. This update brings more fluid, helpful conversations in direct messages too.

Performance & Stability Improvements

Faster Assistant Editor with Large File Sets

  • The Assistant edit page now loads quickly, even when syncing thousands of documents. No more sluggish load times.

Improved Reliability in Model Context Protocol (MCP)

  • We’ve fixed several issues where certain tools weren’t executing properly, and added a loading indicator so you can now see which tool your agent is currently using. That said, some tools may still run into problems when given the wrong input format. We’re aware of this and are actively working on a fix — it’s a top priority and should be resolved soon.

Slack AI App Improvements

  • Fixed a bug where agents didn’t respond correctly when “Reply as a new message” was enabled.
  • Reminder: If you haven’t already, you can reconnect your custom Slack bot to the Slack AI App to place the agent in Slack’s top-right corner for quick access. Check your setup at Settings → Integrations → Slack

Agent Memory, AI Slack App, and More!

Your Agent Now Has a Memory!

Runbear just got smarter—it now remembers your specific instructions and adapts accordingly. For example, tell it "Be more polite," "Always summarize in bullet points," or even "Always fetch pricing details from the web," and it'll personalize future conversations based on what it has learned about you. Give it a try!

Note: This feature is available by default on Claude and still improving. Memory management tools aren't available yet, but your assistant is always learning behind the scenes. Stay tuned—more updates coming soon!

Switch to the New Slack AI App

We’re excited to launch Slack AI App support for custom bots! You can now add AI agents directly to your Slack’s top navigation bar — the fastest way to bring your assistant closer to your team.

Follow this guide to enable the Slack AI App and make it even easier to get the most out of Runbear!

Smart Reply Now Supports Targeted Responses

Smart Reply now gives you more control — you can choose exactly which users or groups should receive the AI Agent's response.

For instance, if you'd like your customer service team to review customer inquiries before replying to the customer, just set them as the recipients. It's an easy way to keep things efficient and accurate.

This update gives you more flexibility to tailor Smart Reply to your team’s needs and workflows.

Track Changes to Your Agent’s Instructions

You can now view and compare changes to your assistant’s instructions over time.

Whether you’re tracking tweaks or debugging behavior, this gives you better visibility into your assistant’s evolution.

Other Changes

  • Support GPT 4.1
  • Improved Model Context Protocol (MCP)
  • Stabilized KB sync

MCP Support, Smart Reply, and More!

April’s here—time to turn the page and start strong! No jokes here—we've genuinely released some incredible new features. We're thrilled to share these exciting updates and preview the important milestones on our horizon!

Model Context Protocol (MCP) Support

Model Context Protocol (MCP) is revolutionizing the AI landscape! We're thrilled to announce that Runbear now supports MCP integrations. This breakthrough transforms your AI agents from simple conversational tools into powerful autonomous systems capable of taking meaningful actions in real-time.

Check out this use case to learn more: Slack MCP Client — Run Claude with MCP directly in Slack

Smart Reply: Your AI Agent Always On Duty

Smart Reply makes your AI Agent feel like a true teammate—always ready to jump in with helpful suggestions without needing to be explicitly mentioned. When someone posts a message in Slack that your agent can help with, it quietly suggests a response just for you. You can then choose to share it with the thread or dismiss it with a click.

No commands, no tags—just seamless intelligence embedded in your team’s workflow.

Use cases

  • Sales team: “Does anyone know the contact at Acme?” → Agent suggests the correct contact info pulled from your CRM
  • HR team: “Did our PTO policy change this month?” → Agent links to the updated Notion doc
  • IT team: “Where’s the VPN setup guide again?” → Agent surfaces the Confluence page instantly
  • Support team: “Who should we escalate this to?” → Agent provides escalation steps based on your internal documentation

How to set up

Smart Reply is automatically enabled for all newly created AI Agents. For existing Agents, you can enable Smart Reply in the Channels > Trigger tab within the Runbear dashboard.

To learn how to set up and use Smart Reply effectively, check out our guide: Building AI Teammates That Actively Join Conversations

Smoother Knowledge Source Integration

We’ve polished the knowledge source integration experience to be faster, more seamless, and more reliable.

Now, you don’t have to wait for all your files to finish uploading when creating a new agent—just keep going, and we’ll handle the rest in the background.

We’ve also added Notion document search during sync, so it’s easier than ever to find exactly what your agent needs.

Smarter Agents, Now Easier to Test

AI Agents Responding to User Mentions

Enable your AI agent to jump in with answers when someone is mentioned in a question—if it has the relevant knowledge to help. Don't worry; only the people involved in the conversation will see the response, privately. Want to share it with everyone? Just hit the "Share to thread" button!

To enable this feature, go to your assistant's Channels tab and toggle on "Trigger on User Mention".

Playground

We’re excited to launch the Playground! Now you can test your AI agents directly in your web browser—no more switching back and forth between tools. It’s all in one place for maximum convenience.