Changelog

🗂️ 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 Has Landed

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.

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!

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

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
  • Runbear now supports OpenAI’s latest GPT-4.1 model, with better instruction following and support for long conversations—up to 1 million tokens of context. More accurate, more helpful responses, even in complex threads.
• Improved Model Context Protocol (MCP)
  • We’ve added a confirmation prompt system for actions that could modify external resources—like sending emails, posting in Slack, or creating Jira tickets. Whenever possible, your assistant will ask for your go-ahead before taking action, helping prevent surprises while keeping things smooth.
  • We also fixed a token limit issue affecting long messages in MCP workflows.
• Stabilized KB sync
  • Fixed issues where some documents were embedded multiple times, and ensured that disconnected files are properly removed from the vector store.

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.

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.

Removed clauses in the Privacy Policy related to sharing user information with business partners and other users from December 30th, 2024.

Improved Knowledge Base Integration Flow

We’ve made syncing with Google Drive and Confluence easier!

For Google Drive, setup is now much simpler. Instead of creating a service account and sharing files manually, you can authorize the integration and select files to sync. Existing setups will still work, but switching requires disconnecting and reconnecting.

Note: Folder sync isn’t supported with the new method. Use the service account option if you need to sync entire folders.

For Confluence, no more manual API keys or URLs—just authorize access and select the pages to sync. It’s faster and more straightforward.

For Slack, syncing is now even more flexible! You can set a starting point for message sync, so only conversations from that point onward will be included. Plus, you’re no longer limited to the default Runbear bot—custom bots are now supported for integration.

Other changes

• Improved Slack message summarization quality: Enhanced the assistant's conversation understanding and summary generation capabilities
• Fixed annotation mapping: Fixed failed annotation mappings in assistant responses

Introducing Perplexity Integration for Real-Time Search

We’re excited to announce that Runbear now supports the Perplexity API, bringing real-time search capabilities directly to your workspace. This integration allows you to summon assistants built with Perplexity to perform instant searches, seamlessly enhancing your productivity within your existing workflow.

Key use cases:

• Market Research: Access the latest market trends, competitor analyses, and industry insights without leaving your workspace.
• Customer Research for Sales Teams: Quickly gather up-to-date information on potential clients to tailor your sales strategies effectively.

To get started, simply select the Perplexity type and enter your Perplexity API key when creating a new assistant. We’re always here to help, so feel free to reach out if you have any questions or need assistance.

Simpler Confluence Integration with Easy Navigation

We’ve enhanced the Confluence page integration experience to make it more intuitive and user-friendly. The new tree view interface allows users to easily explore the structure of an entire Confluence space and select specific pages or subpages for integration with just a few clicks.

Other changes

• Embedded images: We've launched the embedded image feature! This enhancement allows images to be seamlessly included within messages, providing more context and ensuring you stay on track while chatting with Runbear.
• Faster Notion Loading: We’ve made the Notion tree view load faster! Explore your changelog more smoothly and efficiently than before.

Using Slack as a Knowledge Base

We're excited to introduce a new way to leverage your team's Slack conversations. This integration enables your AI assistant to answer recurring questions by learning from your team's historical discussions.

What's New

• Transform your Slack channels into a searchable knowledge base
• Let your AI assistant handle repeated questions automatically
• Save your team's time by reducing redundant explanations
• Keep knowledge base fresh with automatic conversation syncing

Comparison with Message Fetch Action
While Message Fetch is perfect for daily recaps or accessing recent messages (last 30 days), the new Knowledge Base integration builds a comprehensive, searchable archive of your team's knowledge.

Check out our guide for detailed setup instructions.

Annotations in Generated Answers

We’ve enhanced our answer generation with Annotations, which includes references to the knowledge sources used. This feature helps users trace back to the original information sources, making the answers more transparent and reliable.

To enable annotations, go to the Assistant page, select the Channel tab, and open the Preferences section. For a step-by-step walkthrough, follow this interactive guide.

LaserAway Success Story

LaserAway, a leader in aesthetic services, sought a streamlined way to provide timely support for its growing team. By implementing AI assistants in Slack using Runbear, they automated responses to common inquiries, enhancing internal communication and improving productivity. Check out this case study to learn how LaserAway leveraged AI assistants using Runbear to transform their support experience.

Other changes

• Improved error handling when OpenAI API key has insufficient permissions.
• Streamlined the connection process when importing OpenAI assistants.

Assistant Slack Connection Flow Update

We’ve updated the Slack integration flow, allowing each assistant to connect with a dedicated bot featuring custom names and logos, making connections more intuitive and aligned with user preferences.

What's Changed

• Create Slack chatbots with custom @Names and logo images, enabling each AI assistant to have a unique, dedicated identity in their workspace.
• Enjoy a clear and streamlined user experience for the connection workflow.
• Easily manage multiple assistants without any conflicts.

OpenAPI Call Debugging

We've added visibility into API tool calls made by OpenAI Assistants, making it easier to debug function calling integrations.

What's New

• New logging interface for tracking API tool calls within conversation threads
• Detailed context for each tool call execution and its results

How to Access

1. Go to the Assistants page
2. Navigate to the 'Threads' tab
3. Click on a specific thread
4. Click the tool call logs icon to view the execution logs

This enhancement significantly improves the debugging experience when using OpenAI Assistants with external tool integrations.

Other changes

• Improved document loading speed during Notion integration.
• Fixed an issue where Slack Search action couldn't read formatted bot messages.
• Fixed a bug where image interpretation setting would automatically re-enable itself.

Notion Knowledge Base is Now Available

We're excited to announce that Notion Sync is now available in beta. This allows you to keep your OpenAI Assistant up-to-date with the latest changes in your Notion documents automatically. No more manual uploads needed to ensure your assistant uses the most current information.

To synchronize is super simple:

1. Connect your Notion workspace to Runbear
2. Select the pages you want to sync
3. Runbear will automatically upload the content to the OpenAI platform

Tip: Initially loading all documents can be time-consuming. We recommend starting with only essential pages for a quicker setup.

Organization-wide Analytics Dashboard

We're thrilled to launch our Organization-wide Analytics Dashboard. This new feature allows you to monitor the overall performance and usage of AI assistants across your entire organization.

While starting with these functionalities, we plan to expand the dashboard with more comprehensive analytics and insights in the future, including token usage data and more detailed performance metrics.

Renewed Slack Search

We've made significant improvements to our Slack Search functionality based on user feedback, making it more intuitive and user-friendly:

• Default Behavior Change: The Assistant now automatically searches the history of the channel where the query is made, rather than pre-selected channels.
• Expanded Search Capability: To search other channels, users can now simply tag the desired channel names in their query (make sure the Slack bot is invited to these channels).

How it works:

1. Ask a question in any channel where the Assistant is active - it will search that channel's history by default.
2. To include other channels in your search, just tag them in your query (e.g., "What was discussed about project X in #general and #project-y?")

These updates make the Slack Search more intuitive and accessible for all team members, regardless of who set up the Assistant.

Note for existing users: Your current setup will continue to work as before. To benefit from these new features, simply update your Assistant.

We're continuously working to improve Runbear based on your feedback. As always, we welcome your suggestions for further enhancements.

Other Changes

• We fixed an issue where some Google Drive docx files were not uploading properly.
• We resolved a problem where Custom Slack Bots weren't automatically triggering in threads.

Slack Conversation Search

We're excited to introduce our new Slack Conversation Search feature, empowering your AI Assistant with the ability to retrieve and analyze conversation history from specific Slack channels.

Getting Started:

1. Enable the Slack Search action for your Assistant.
2. Select the Slack channels you want your Assistant to have access to.
3. Invite the Slack bot to the selected channels.
4. Start asking your Assistant questions about past conversations, decisions, or updates.

Key Benefits:

• Quickly catch up on missed discussions
• Track project progress and milestones
• Recall important decisions without digging through conversation history

It's crucial to understand that this feature is distinct from Slack as a knowledge base functionality. The Assistant doesn't incorporate or store Slack conversation history as part of its knowledge base. Instead, it performs targeted, real-time searches of specified Slack channels only when you ask a specific question.

This approach ensures that the Assistant accesses only the most relevant information when needed. For this reason, we recommend connecting only the essential channels that contain the information you're likely to query.

Other Changes

• Resolved synchronization issues with Confluence knowledge base, ensuring more reliable data updates.
• Fixed an issue where Runbear occasionally did not post comments when tickets were created in Zendesk.

Introducing Custom Bot

We're thrilled to announce the launch of Slack and MS Teams custom bots! You can now create multiple AI assistants under your own brand across your Slack and MS Teams workspaces. This new capability allows you to build dedicated bots for specific purposes, enhancing your workflow and user experience.

You can build a custom bot by accessing the Integrations > Slack or Teams page.

Thank you to our beta testers for their invaluable input in shaping this feature. We can't wait to see how you'll use it.

Feedback Dashboard

We're excited to launch our Assistant Analytics Dashboard in beta. This new feature allows you to monitor the quality of your Assistant's answers with emoji-based feedback tracking.

this dashboard currently focuses on tracking response quality. While starting with this functionality, we plan to expand it with more comprehensive analytics and insights in the future.

Code Interpreter Support

You can now enable the Code Interpreter tool when creating an OpenAI Assistant. Code Interpreter allows your assistant to process various data files, generate graphs, and tackle intricate math problems through iterative problem-solving.

When enabled, any files generated by Code Interpreter during Slack conversations will be automatically uploaded to the relevant Slack thread.

New triggers for MS Teams & rebranding

We're excited to bring many of the triggers and preferences previously exclusive to Slack to Teams! Now, Teams users can enjoy even more flexibility and customization. (You need to reinstall the Runbear app to access new features, if you have it installed.)

Along with our recent rebranding, the Teams app name has changed from PlugBear to Runbear. Be sure to mention @Runbear when you need assistance!

Other Changes

• Enhanced image interpretation with native vision.
• Improved response accuracy in threads with file attachments.
• Boosted stability when uploading large files.
• Bug fixes:
  • Long responses now send correctly.
  • Better stability when using the MS Teams app.

Spreadsheet Support for Knowledge Sources

We've expanded our knowledge source capabilities! You can now upload XLS and CSV files to OpenAI assistants, making it easier to incorporate structured data into your AI workflows.

Revamp UX

We've given our web app a major facelift to improve navigation and prepare for upcoming features:

• Distinct look & feel from our landing page, with a more organized layout
• "LLM Apps" are now called "Assistants"
• Connections tab moved to assistant detail pages
• Channels tab renamed to "Integrations" We'd love to hear your feedback on these changes! If you notice anything that feels off or have suggestions for improvement, please let us know. Your input is crucial as we continue to refine the user experience.

Google Gemini Now Available 🎉

We're excited to announce that Runbear now supports Google Gemini! You can now create Gemini assistants with the following models:

• Gemini 1.5 pro
• Gemini 1.5 flash
• Gemini 1.0 pro

Our current Gemini integration focuses on text generation, with plans to add multimodal support in the future. We welcome your thoughts and experiences with Gemini as we continue to evolve in the AI landscape.

Other Changes

• Custom GPT Import Fix: We've restored the functionality of our bookmarklet for importing Custom GPTs. You can now successfully download and upload files attached to Custom GPTs when importing them to OpenAI Assistants.

URL Fetch - Empowering AI with Web Page Access

We’ve introduced the “URL Fetch” feature to supercharge your AI capabilities 🎉

By enabling your AI to read web pages, you can unlock powerful new tools—whether it’s summarizing content, creating marketing posts inspired by existing examples, or extracting metadata from news articles.

To activate this feature, head to the “LLM App Edit” page, expand the “Advanced Options” section, and switch on “URL Fetch.”

Check out the “AI-Powered Web Page Summarization for Slack” use case to learn how it works!

New Use Cases Library - Your Inspiration Hub

Curious to discover more use cases? Explore our new and expanding Use Cases page! We’re adding fresh examples every week, so stay tuned for more insights.

Other Changes

• Shared Google Drive Sync: Now supporting Google Shared Drives. Enable by adding the service account to your Shared Drive team. Expand your AI's knowledge base with team-wide resources!
• Improved Auto App Selection: Better matching of user queries to the most relevant AI app in multi-app channels.

Runbear is SOC 2 Type II Compliant!

We’re excited to share that after months of rigorous preparation and auditing, Runbear is now SOC 2 Type II certified. This means we meet high standards for data security and privacy, ensuring your data is safe from unauthorized access and breaches. Achieving this certification shows our ongoing commitment to protecting your information and providing a secure service.

Thank you for your continued trust and support. For more details, check out our blog post.

Public Roadmap is Now Live!

While we regularly share recent updates in this changelog, you might wonder what’s coming next. We’ve been hearing from our customers via various channels, but as we believe in the “build in public” approach, we want to share our product roadmap and hear more from you.

Our public roadmap is accessible via Runbear Roadmap and through the footer on our main website. Drop by to see what’s coming up, vote for the features you’re excited about, and suggest new ones.

Please note that all features in the roadmap are subject to change, especially those with unspecified or longer timelines. Your votes signal us that a feature is high priority.

Other Changes

Fixed an issue preventing the synchronization of large Google Drive files with OpenAI Assistants.

PlugBear is Now Runbear

We are excited to announce that PlugBear will be officially renamed to Runbear this coming weekend! This change reflects our commitment to a stronger brand identity and better alignment with our company’s vision. All existing app configurations will continue to work seamlessly, and you will have the option to switch to the new @Runbear mention. For more details on the transition and what to expect, please read our full announcement.

Image Generation with Dall-E

Now you can enable DALL-E image generation for the OpenAI Assistants app connected to your Slack channel. It can generate images using DALL-E based on your chat messages and prompts. You can enable DALL-E image generation by simply enabling the “DALL-E Image Generation” option in your app.

Other Changes

• Improved Google Drive Sync: Resolved an issue preventing the synchronization of large Google Docs and Slides files. Now, all file sizes can be synced seamlessly.
• Trigger on Channel Message: We’ve added an option for the LLM App to respond to every new message within a Slack channel.

Claude 3.5 Sonnet Support

We're excited to announce that we now support Claude 3.5 Sonnet, a state-of-the-art AI model with exceptional performance and cost efficiency. Claude 3.5 Sonnet excels in several key areas, including context precision, visual reasoning, and coding tasks, surpassing models like GPT-4o, Gemini 1.5 Pro, and Llama 3 400B in specific benchmarks.

You can select the Claude 3.5 Sonnet model when creating a new LLM application on Runbear. Currently, Runbear’s integration with Claude 3.5 is focused on text generation. If you need additional support or have feedback, please feel free to contact us.

Current Date Awareness

OpenAI Assistant within Runbear now has the capability to access the current date. This means it can answer your date-related queries accurately, like “How many visitors did we get in the past week?”. This enhancement allows the Assistant to seamlessly incorporate current dates into its responses, improving the overall user experience with time-related questions.

Use Case

By integrating with databases or search APIs, you can now ask natural questions such as:

1. Visitor Analytics: Easily query the Assistant with questions like “How many visitors did we get in the past week?” to quickly obtain accurate visitor data, enhancing your ability to monitor and analyze traffic trends.
2. Sales Tracking: “What were our sales figures last month?” to get precise and timely updates without manual date calculations.
3. News Summarization: “Summarize today’s news” to receive a concise overview of the latest headlines and events.

Custom-Branded Slack App

We're thrilled to announce that our Slack Branded Bot feature is now available in private beta. This powerful new capability allows you to create custom-branded Slack apps and manage multiple Slack workspaces from a single Runbear organization. To apply for early access and become a design partner, please contact us.

Other Changes

• Reply as New Message: Now, you can choose to have the Assistant's responses sent as new messages directly to the Slack channel, rather than just within existing threads. This allows you to integrate Runbear's responses with other tasks more flexibly. You can enable this feature by navigating to Connections > Edit > Preferences.
• Big Fixes: We fixed a few minor bugs, including one that disabled the Slack Connection Preferences update button.

Support for Disabling Annotations

We have introduced a new feature that allows you to turn off annotations in the responses of your LLM applications. Previously, all responses included annotations by default, indicating the sources referenced. With this update, you now have the flexibility to turn off annotations when they are not necessary. This enhancement helps streamline the response output and tailor your user experience to specific needs.

You can enable or disable the annotation settings under Connections > Edit > Preferences in the Runbear dashboard.

Other Changes

• Improved the assistant import form: We have improved the form for importing Assistants on the dashboard. The import and post-import editing processes are now more intuitive, making it easier to manage your Assistants.
• Confluence Sync Error Fix: We have resolved an issue where errors occurred during the synchronization of a large number of Confluence pages. The synchronization process is now more stable and reliable.

Confluence Sync Now Available

Following our recent update with Google Drive Sync, we're excited to announce that Confluence Sync is now available. This allows you to keep your OpenAI Assistant up-to-date with the latest changes in your Confluence documents automatically. No more manual uploads needed to ensure your assistant uses the most current information.

To synchronize is super simple. Connect your Confluence space to Runbear and add the pages you want.

We're not stopping here. We plan to add more data source integrations and would love to hear your suggestions. Let us know your preferred platforms and knowledge bases, and we'll work on bringing those to you next.

Use Cases

Build your assistant chatbot using confluence. It helps you manage knowledge base up-to-date without hassle.

You can create different chatbots in each channel, for example:

• A company policy chatbot in #hr-qna channel
• A compliance assistant in #compliance channel
• A customer support assistant in HubSpot or Zendesk that generates draft
• A brand guide assistant that checks if creatives meet the company requirements

and more!

Other Changes

• Enhanced Auto App Selection: Runbear now makes smarter app selections. If you have multiple apps connected to one channel, it can better determine which LLM app to invoke using a broader range of thread contexts.
• Improved Emoji Triggers: Emoji triggers are now activated regardless of the emoji skin type.

This week, we mainly focused on overall feature stabilization, including improvements to Google Drive sync, GPT import, and Slack responses. A big thank you to all our customers who have reported issues and patiently borne with us during this process.

More Controls Over OpenAI Assistants

You can now control the maximum completion and prompt tokens to manage token usage, which affects your OpenAI costs. Adjusting the maximum prompt tokens helps control the context window of your assistant. For longer threads, limiting the maximum prompt tokens will help save costs by truncating the context for more recent messages, though it may miss some of the oldest contexts.

You can also control your assistant’s temperature or top_p values to manage the consistency and randomness of the output. You can update these settings in the Assistants.

Support for Slack Trigger Based on Legacy Slack Bots

We noticed our ‘Auto-Trigger by User or Bot’ feature couldn’t be set up with some legacy Slack apps, preventing you from setting up triggers on those apps. We added an alternative way to integrate, and you will be able to find the “Legacy Bot” tab in the trigger setup.

Other Changes

• We made the gpt-4o model the default suggestion, as it outperforms other models with faster generation and is half the cost. We encourage migrating to gpt-4o if you’re using gpt-4 or gpt-4-turbo.
• Google Drive Sync: Added support for Google Slide files and stabilized periodic synchronization.
• OpenAI Assistants: Fixed intermittent generation failure when annotations are created.
• Integration for Slack: Fixed a bug with ephemeral responses.

The world is changing fast, and we're excited to be part of it! With OpenAI's new model, GPT-4o, you can now use Runbear for faster and more powerful AI.

Even better, you can use your Google Drive as a knowledge base. This means you can keep your information updated without needing to upload new files.

Enjoy the future of AI with Runbear!

Dynamic Knowledge Source with Google Drive

Keeping your Q&A bot's answers current just got easier. Say goodbye to manually updating your knowledge sources.

With our new Google Drive sync, your OpenAI Assistant can now automatically keep its file search aligned with your chosen Google Docs. This means your assistant will always draw from the latest information in your documents, without the need for repeated file uploads.

Learn more about Google Drive sync!

GPT-4o Support

You can now use OpenAI's new model GPT-4o using Runbear! GPT-4o is 2x faster, half the price, and has 5x higher rate limits compared to GPT-4 Turbo. It also better handles non-English languages and uses a more efficient tokenizer.

To use GPT-4o, simply visit your Assistants, click the edit button, and switch the model to GPT-4o.

Other Changes

• Bump up bookmarklet: We have upgraded our bookmarklet to version 3.1, which includes a critical fix for an issue where imports were failing in certain accounts due to recent updates to GPTs.
• OpenAPI Add-on: We have added support for authentication based on query strings and fixed a bug related to tool argument parsing.

In closing, We sincerely thank everyone who supported Runbear by upvoting us on Product Hunt last week. Your engagement is invaluable to us, and we're grateful for your enthusiasm and feedback as we continue to improve. Thank you for being part of our journey!

Image Interpretation #slack

Runbear now recognizes images on Slack! Just send over an image with what you need to be interpreted or extracted in the chat. You can use this feature for the following purposes:

For Marketing Teams

• Creative image compliance check bot
• Creative image feedback bot to enhance campaign performance

For Product Teams

• UI Design review bot

For Customer Service Teams

• Customer support bot that responds to inquiries with images

Update your OpenAI Assistant directly in Runbear

Updating your OpenAI Assistant's instructions and files is now simpler. Make your changes directly in Runbear, and they'll automatically synchronize with the OpenAI platform—no need to manage your LLM apps elsewhere.

Other Changes

• Thread Context Support in Slack Group DM: Runbear is now able to understand and respond to thread context in Group DM channels the same way it does in public channels.

In closing, our Slack app is now officially approved and listed in the Slack App Directory! A big thank you to our current customers for your early trust and support. This update doesn't impact those who have already installed our app, but it's a step forward we wanted to celebrate with you!

OpenAI Assistants v2 Support

OpenAI Assistants in Runbear now operate on v2. All existing LLM apps based on Assistants v1 have been automatically upgraded to v2. The new version allows you to upload up to 10,000 files, in contrast to the previous limit of 20, providing your assistants with more context. Be aware that your OpenAI bill will be charged based on the size of the uploaded files at a rate of $0.10/GB/day.

Annotation Support in Slack

Our latest update introduces annotation capabilities for responses generated by the LLM App within Slack. This promotes transparency by allowing users to easily reference and verify the original source or document related to the app's responses.

Use Case Spotlight: Building Your SQL Assistant

Imagine having an AI at your fingertips that crafts SQL queries based on your specific database schema. This SQL Assistant not only generates the queries for analyzing your data but collaborates with you to refine and perfect them, making your analytics tasks more efficient and effective. Learn how to create your own SQL Assistant!

Other Changes

• When you add the Runbear bot to a channel you will receive a helpful welcome message guiding your next steps.
• Bug fix: Resolved ephemeral response setting issue for Slack connection

Custom message template

Introducing Custom Message Templates!
Want to enhance the context of your messages sent to LLM? We've got you covered with our latest feature. Now, you can specify who sent the message or which thread you're addressing. Simply head to the Connection Edit page and navigate to the Preferences tab to create your personalized message template. Tailor your messages to your liking and make every interaction count! Here is an example: Type like this. Then ask, "Tell me my user id," and you'll receive your user ID!

Simplifying LLM App Creation

This week, We've made enhancements to streamline the creation of LLM Apps. Here’s what’s changed:

• Ready-to-Use Templates: To help you get started immediately, we've introduced a range of pre-built templates. These templates are designed to cover various common use cases, enabling you to quickly deploy customized AI solutions without starting from scratch.
• Simplified App Form: We've redesigned the App creation form, making it more intuitive and user-friendly.

Streaming Answers

Great news! 🎉 If you're using OpenAI Assistants, you no longer have to wait until the entire answer is generated. Runbear now streams the answers, allowing you to view them as they're being generated.

Enjoy your quicker AI responses!

Use Cases

Daily News Summarizer

Create a personalized bot that summarizes daily news on your chosen topics, delivering concise and relevant updates directly to your communication channels like Slack or Microsoft Teams. Check our blog post to learn more!

Answering Questions Before You Do

You can create a bot for Slack that can answer questions on your behalf. People can still ask questions the same way they always have – by mentioning you or your team.

By allowing a bot to handle routine questions, your team members can focus on the important work that truly matters, enabling you to achieve more.

Beta Support for Anthropic Claude 3

The early March release of Claude 3 was significant, and many customers have expressed interest in trying it. Now, you can head over to Runbear and create your Claude 3 app to start testing right away.

While the Opus model is the most sophisticated model, be mindful of its higher cost. Currently, Runbear’s integration with Claude 3 is focused on text generation(with vision support planned for the future). We invite everyone interested in Claude 3 to dive in, experiment, and please share your feedback with us.

Use cases

Customizable Customer Support AI Agent

Generative AI is changing the game for customer support, offering immediate access to a company's internal knowledge base. Imagine AI assistants that summarize inquiries, draft suggested answers, possess the knowledge to resolve issues, and offer much more!

Do you want your AI assistants in Zendesk or HubSpot? Check out the blog post for more details!

Custom AI for Customer Research

Are you looking to quickly research and qualify your leads? With Runbear, you can build an AI to do just that. Enable your LLM applications to automatically share summaries of customer interactions—whether they're signing up for your service or asking questions through HubSpot. Learn more about AI for customer research!

Other Changes

• Enhanced MS Teams Integration: Addressing user feedback, we've streamlined the integration process with MS Teams to reduce potential difficulties.
• Option to Exclude Runbear Links in Error Messages: This update eliminates Runbear-directed error description links, optimizing the experience for users handling customer interactions via Runbear.

Zendesk Integration: Enhance Your Customer Support Experience!

We’re excited to announce a new channel in this update: Zendesk! Runbear now supports customer service agents to work faster and more efficiently than ever before.

How does it work?

When a ticket is created, Runbear understands its context and responds with an internal note that only the agents can see.

Use cases

Automated Email Drafts: Automatically generates email drafts based on Zendesk tickets, tailored to the customer's inquiry. This saves CS agents valuable time and enhances customer interaction quality.

Inquiry Resolution Guide: Upon ticket creation, Runbear utilizes the knowledge base to perform preliminary investigations, offering insights and potential solutions to agents. This approach enables agents to quickly identify potential resolutions and enhance their response to customer inquiries.

Other Changes

• Refined Task Status Indicators in HubSpot: Runbear now provides more apparent status updates for tasks within HubSpot, detailing reasons for any delays or errors. In request failures or delays, users can quickly identify the precise cause and the corrective actions required.
• Usability Improvements: We've made a few improvements to enhance user experience, including adding a search function for selecting LLM Apps.

Microsoft Teams Integration Now Available 🎉

We are pleased to announce that Microsoft Teams is now available. Our application has successfully passed the Microsoft App Verification process and is integrated with Microsoft Teams as a beta feature. We invite you to explore this new functionality and welcome your feedback!

Use cases

The Microsoft Teams integration accommodates the majority of use cases that Slack supports. For further details, please refer to our usecase docs. Here is one of the use cases, a proof-of-concept:

Trigger By Multiple Keywords #slack

Now, there's no need to create new connections for each keyword. You can configure a single connection to recognize multiple keywords, simplifying your setup.

For instance, you can trigger the Benefits Q&A bot with questions containing keywords such as "health benefits", "medical coverage" or "wellness programs".

HubSpot Integration Now Available 🎉

We're thrilled to unveil our new HubSpot app integration with Runbear, offering a streamlined approach to enhance ticket resolution speed and agent productivity. This integration allows for seamless ticket management within HubSpot, simplifying the workflow for customer service teams.

Use cases

Automated Ticket Investigation: Upon ticket creation, Runbear leverages its knowledge base to conduct preliminary investigations, offering insights and potential solutions to agents.

Draft Response Generation: Runbear can automatically generate a personalized email draft whenever a ticket is created. This enables your team to greatly streamline the email response process and improve the customer experience overall.

We are eager to support a wider range of use cases within HubSpot. If you have any specific use cases or ideas you'd like to see implemented, please feel free to share them with us!

Custom App Icons (beta)

You can now customize app icons! To do so, visit the Connections page and select a connection. Then, open the Preferences tab to begin customizing icons. This beta feature will be available for all plans until its official release. (Disclaimer: This feature only customizes icons of generated messages. Stay tuned for more advanced customization options!)

This week, we're thrilled to introduce significant updates to Runbear, enhancing your experience with new features and improved functionalities. Here's what's new:

Call External API From OpenAI Assistant, Using Open API

We've enhanced our OpenAI Assistants app to support OpenAPI, a significant improvement that simplifies and strengthens the process of connecting OpenAI Assistants with external APIs. It enables your Assistants to interact more effectively with the external world.

For instance, you can extract information from user queries and save them into Airtable, activate Zapier Actions for extended functionalities, and more! The range of possibilities is now even wider.

Our integration guide provides detailed setup instructions. Stay tuned for our upcoming guide on specific use cases.

OpenAI Assistants Request History

Nobody likes seeing "Oops" messages, especially when figuring out why they're happening is challenging. We're excited to tell you that you can now see your OpenAI Assistants' past requests on the connections page. Just click on a connection on the Connections page or use the link in the error message to see the information.

Connection Preferences

Customize your connection preferences to suit your needs! For instance, you can opt to receive generated answers through an ephemeral message that only you can see. Additionally, you can choose to use a single message under a thread to prevent interference or optimize costs. You can find these preferences on your connection page. More options are coming soon, so stay tuned!

Conditional Trigger by User or Bot 🤖 #slack

This new feature automatically triggers responses based on messages from specific users or bots within Slack. So, you can say goodbye to the hassle of repeatedly mentioning @Runbear.

Use Cases

On-call Assistant: Set up an automatic trigger for the messages by an incident alert bot to propose initial response strategies when an outage alert is issued.

CS Assistant: Configure triggers for customer service notification bots, such as Zendesk, to present response manuals to customer service representatives when a ticket is created.

How to Setup?

You can find the option under the "Triggers" section on the Connection editing page.

Other Changes

• UI Improvements: We've improved the Connection page more user-friendly and intuitive
• Bug fix: Resolved a bug where unintended Connections were selected for keyword triggers.

Support Conditional Trigger by Emoji 🐥 #slack

Looking for a simpler way to interact with Runbear? Say hello to our emoji-trigger! 🎉 Now, with just an emoji, you can invoke Runbear to respond to individual messages within your conversations. This means that Runbear will act based on the specific message you've marked with an emoji. Here's a simple use case where a customer support agent swiftly gets assistance on a customer query with a simple emoji.

Other Changes

• Enhanced Messaging for Slack: We’ve improved how long messages are handled in Slack. Now, if a message from LLM is too lengthy, you'll see a "See more" button to expand the content fully. No more cut-offs!
• Slack Attachment Handling: With the recent introduction of thread-level file upload and knowledge retrieval, we've now stabilized file handling and ironed out the edge cases. It’s still early days for this feature, so we welcome all your feedback to make it even better.

Update on Microsoft Teams Integration

We know many of you are eagerly awaiting the integration with Microsoft Teams. We appreciate your patience! We’re working closely with Microsoft and just waiting on the final nod from them, so stay tuned!

Runbear launches Conditional Triggers, Attachments Support, and new use cases.

Attachments Support in Runbear #slack 🎉

Runbear now allows the processing of uploaded files. Easily upload documents in PDF, docx, or plain text formats and interact with your LLM apps to analyze the content. Ask your LLM apps to process and respond to queries related to the uploaded files.

Use Case: Document Review Bot

Reviewing documents to extract crucial information can be a daunting task, requiring significant time and effort. Visit Document Review Bot to learn more about how to build the LLM application.

Conditional Triggers for LLM Apps #slack

Your LLM apps can now be set to interact with messages based on defined conditions, not just when @Runbear is mentioned. This enables a more natural experience in chats where the LLM apps can provide input whenever a message contains a specific keyword.

For example:

• When a message contains "error reported," the app can explain the error and suggest a solution.
• When a message contains "new customer," the app can respond with the customer information.

More Upcoming Trigger Conditions

We are diversifying trigger conditions to support a wider range of use cases. Soon, you will be able to call the app under a variety of conditions as follows:

• Emoji-based trigger: Attach an emoji to a message to trigger the app.
• Scheduled trigger: Schedule the app to run at a specific time.

Today, we're thrilled to unveil the first in our series of advanced chat triggers. With these new features, we aim to empower you to create even more innovative and efficient workflows.

Thread Auto-Reply

Introducing the Thread Auto-Reply trigger, designed to streamline your conversational experience effortlessly. Engage with your selected LLM App in a thread as you would with ChatGPT, without needing to mention @Runbear. This feature can be set per app connection, paving the way for smoother interactions.

Stay tuned for upcoming automatic trigger conditions, including emoji (😄) reaction triggers and responses triggered by predefined criteria such as keywords or user mentions.

Other changes

We've resolved an issue where chats could intermittently become unresponsive due to connectivity problems between Runbear and the OpenAI Assistants API, which sometimes led to hangs in the request. This fix significantly enhances the reliability of chat responses.

Upcoming Use Cases

We are developing features to support more use cases, including:

• Automatic Translator: It will translate every message in a channel into other languages.
• Daily Digest: It will summarize daily updates made the day before.

...and more! Stay tuned.

Last Limited-Time Offer 🎁

It’s my bad. Last week, we didn’t share that the offer ends today.

Since the launch of Runbear this month, we have received an overwhelming number of subscriptions from users. We are incredibly grateful for your support. Thank you!

As a token of our gratitude, we have decided to extend the limited-time offer for one more week. Make sure not to miss this opportunity! The offer will remain valid until next Wednesday.

Web Browsing (beta)

Your OpenAI Assistant can now browse the web! 🔍🌐 We are thrilled to announce that your OpenAI Assistants have the ability to access a web browser, allowing them to utilize vast amounts of information. With this new capability, you can create even more advanced Assistants, like the Sales Copy Generator, which can leverage prospects' online search results. Check out Web Browsing for more details.

Use Case: Sales Copy Generator

Enhance your outreach with a customized sales copy generator. Personalized messages grab attention, but writing each one is slow. With OpenAI and Runbear, you can create your Sales Copy Generator in under an hour.

Runbear Initiates SOC2 Compliance with Vanta 🔒

Runbear is proud to announce the start of our journey to SOC2 compliance, partnering with the security and compliance leader, Vanta. This initiative underscores our unwavering commitment to maintaining the highest standards of data security and privacy for our users. Stay tuned for updates as we reinforce our promise to deliver a secure and trustworthy platform. Check our Trust Center to learn more!

Loading Indicators

Runbear now sends a status message right after accepting your requests. This update provides continuous status updates to keep you informed about the operational status of the app or any retry efforts.

Editing Connections

Finally, you can now edit connections! We have made connections editable before adding more controls. Enjoy the ability to create and edit connections between channels and LLM apps.

Use Case:

We have outlined the process of creating your Q&A bot using Notion pages. Additionally, you can create and import the Q&A bot using OpenAI GPTs. Take a look at the Building a Q&A Bot guide to build custom Q&A bots tailored to your specific requirements.

More Upcoming Channels

Microsoft Teams integration is coming soon! (We are waiting for Microsoft's review.)

In addition, we are prioritizing services to support integration. Please take a moment to complete our 1-minute survey to help us prioritize the most critical services for shipping!

Slash Command to call your app

Elevate your Slack experience! The new slash command /plugbear ask YOUR QUESTION simplifies interactions with the LLM app—making it fast, straightforward, and efficient.

Commitment to Security

Our dedication to robust security is unwavering. Currently undergoing SOC2 Type 2 compliance auditing, we are committed to ensuring that our security infrastructure adheres to the highest industry standards.

Chat System Enhancements

We have upgraded our messaging system to provide a more efficient handling of chat requests. While we acknowledge that some users may have experienced temporary delays, we are dedicated to continuously improving the system for a seamless experience. We appreciate your understanding and apologize for any inconvenience.

Upcoming Microsoft Teams Integration

Exciting news! The integration with Microsoft Teams is on the horizon. Our application is in the review process, and we are anticipating approval shortly.

Don't miss our latest resource - the Building Q&A bot use case guide, offering insights on leveraging Runbear for enhanced Q&A interactions in your organization.

Runbear now supports OpenAI GPTs conversion, listing apps in channels, and connecting LLM apps to private Slack channels.

OpenAI GPTs Integration

We have launched a new feature that allows users to import OpenAI GPTs and convert them into OpenAI Assistants. This enhances Runbear's functionality by enabling integration with the latest OpenAI GPTs.

Private Channel Support

In response to user requests, Runbear has expanded its capabilities to include support for private Slack channels. You can now connect your preferred LLM app to private channels, ensuring that sensitive conversations remain exclusive while still benefitting from Runbear's chat-based tool integration.

Listing Connected Apps

Users can now easily manage their apps connected to channels on Slack with the new /plugbear list apps command. This command provides a quick overview of which apps are connected to which channels, streamlining channel app management.

Use Case: Thread Summarizer

Have you ever felt tired of reading a long thread? Let AI summarize the conversation for you. You can build a bot for Slack in just 5 minutes by following this guide: Building a Bot for Summarizing Threads.

Happy New Year, Runbear users! We are excited to announce the official launch of Runbear!

Runbear Docs

We have published a product documentation site that provides a usage guide and introduces use cases. Please refer to the documentation at docs.runbear.io.

Auto App Selection

Managing multiple apps and their Slack channels can be cumbersome. With the Auto App Selection option, you no longer need to assign specific channels to an app to avoid receiving multiple answers. Turn on this option and get the most relevant app's answer.

Retrying on Rate Limit

We have seen you encountered error messages caused by OpenAI's rate limit. To address this, we will handle queueing and retrying requests automatically.

Pricing

The Runbear team has extensively discussed and decided on the best values and pricing options for our users. Today, we are finally publishing our pricing model, which will allow us to deliver more features and value continually.

You can find the supported features and pricing details on the Pricing page.

Early-Bird Discount 🎁

To show our appreciation for your support, we are offering a 50% discount until the end of January 2024. Claim your discount by upgrading your plan on the Billing page.

We cannot express our gratitude and love to you enough. We promise to continue delivering even more value to you.

Enhanced Customization for LLM Apps

We've introduced a description field to LLM apps, which will become even more useful with our upcoming release. Additionally, you can now easily modify existing configurations after creation, simplifying the process of managing your apps.

Improved Error Log Accessibility

Building on our invocation history page, we've increased the accessibility of error logs. When an LLM app-related error, such as an OpenAI rate limit issue, is detected, you'll be notified with a link to the invocation details directly in your Slack or Discord app.

Official Launch Countdown

With just one week to go until the official launch of Runbear, we're gearing up to unveil new features and enhancements that will enhance your Runbear experience.

As we bid farewell to 2023, we're eagerly anticipating the new updates and features coming in 2024. Stay tuned for a more advanced Runbear experience in the year ahead!

Runbear Python SDK Released

We're pleased to introduce Runbear Python SDK. This new addition simplifies the integration of custom LLM apps with Runbear. Now, you can easily integrate your own AI toolkit, such as OpenAI, Anthropic, Gemini, and many others, into Slack or Discord.

Learn how to use the SDK in our guide: Integrate Your Custom LLM App with Runbear.

Enhanced Troubleshooting with Invocation History

Understanding error messages from your Runbear bot just got easier! We've introduced a feature that allows you to view your LLM app’s recent invocation history. Now, you can effectively troubleshoot error responses from LLM apps. Access this feature from the "Connections" detail page.

Other Changes

• We've enhanced the stability of the Slack workspace installation process.

LangServe Integration

LangServe integration is now available on Runbear. You can seamlessly connect your LangChain app, served by LangServe, to Runbear, unlocking a host of new capabilities.

Discover more about this integration in our guide: Integrate Your Custom LangChain Apps with Slack through Runbear.

Enhanced Slack Thread Context

Runbear has improved its context awareness within Slack. It now tracks the entire conversation threads — not just the messages that directly mention Runbear. This results in more relevant and coherent follow-up responses in ongoing discussions.

Other Changes

• Discord integration now supports the calling of API functions, a feature previously exclusive to Slack.
• We've resolved an issue with listing channels in Slack workspaces that have a large number of channels.

API Function Calling

Runbear now supports calling API functions. This feature allows you to call API functions from your LLM app, enabling you to build more complex apps.

Learn more about this feature in the How to Call APIs using OpenAI Assistant guide.

Slack Direct Messages

You can ask Runbear by sending a direct message to it. This is useful for asking questions privately.

Other Changes

• Improved speed
• Added a guide document on converting converting GPTs to Assistant

Runbear was launched on Product Hunt last Thursday and achieved #5 Dev Tool of the Week! 🎉 We sincerely appreciate your support and are thrilled to share our values with you.

Discord integration

Responding to numerous requests, Runbear now offers integration with Discord. This feature is particularly beneficial for organizations, especially open-source projects, allowing them to incorporate LLM apps into their community. We are excited to support the open-source community through this integration!

#channel

You now have the option to select specific channels on Slack and Discord for your LLM app. This flexibility facilitates the use of various apps within a single organization. For instance, you can assign an onboarding Q&A chatbot to the #people-team channel while a data catalog bot operates in the #data-team channel.

We hope you enjoy the versatility of using multiple LLM apps in your workspace 😄