📝 Release Notes

QueryPie AI Hub is continuously evolving to deliver a better AI integration experience. On this page, you can find the key updates and improvements for each release.

2025/07/28 Release

New Features

Widget feature added: Easily save and revisit statistical data you’ve extracted via LLM. Perfect for when you want to keep important insights handy.
Billing feature added: A new billing system has been introduced with three plans: Starter, Team, and Enterprise. Each plan provides a specific amount of credits that can be used within the Organization. You can also view the credit usage within your organization for better management and transparency.
Audit log feature added: An audit log is now available for organizations. Actions performed by organization members are recorded with the event time, user, and event details. Currently, the audit log tracks LLM chat events, settings changes, and MCP tool invocations. More event types will be added in the future for enhanced traceability.
Salesforce MCP added: Execute SOQL/SOSL queries, retrieve object schemas, perform CRUD operations, and run aggregate analytics directly on your Salesforce data through comprehensive database management tools.
QueryPie KAC MCP added: Manage Kubernetes clusters across AWS EKS and on-premises environments with comprehensive tools for resource CRUD operations, metrics monitoring, and event tracking through centralized API protection.

Improvements

Now you can use Kakao Map, Kakao Navigation, and Daum Search MCP without having to set up an API key.
Fixed chat functionality issues including message display errors, and file upload problems to improve user communication experience
Resolved system stability issues including auth token handling, preset validation errors, and tool cache invalidation problems for enhanced performance
Addressed UI/UX issues including organization management bugs, share functionality problems, and Edge Tunnel connection issues for better user experience

2025/07/14 Release

New Features

Edge tunnel: Securely connect AI Hub with your private network. Now, you can link MCPs, databases, servers, QueryPie, and more running in a private network to AI Hub, enabling even greater extensibility.
Organization feature added: You can create organizations and invite other users to manage them together. Additional security features such as Audit Log and DLP will be added to organizations in the future.
Filesystem MCP added: Browse and edit your local file system through the Filesystem MCP. By connecting a tunnel to AI Hub via CLI, you can directly access your local file system from AI Hub.
Context7 MCP added: Context7 MCP allows LLMs to refer to the latest technical documents and code examples, providing more accurate answers.
Airtable MCP added: View and edit Airtable data directly through the Airtable MCP.
Discord OAuth MCP added: In addition to the existing Bot token method, Discord MCP now also supports OAuth authentication.

Improvements

Fixed functional errors in Jira, Confluence, and Notion MCPs
Clearer error messages when tool list loading fails in Custom MCP SSE

2025/06/30 Release

New Features

Smart Tool Discovery: Instead of exposing all MCP tools registered in a preset, only the necessary MCP tools are dynamically searched and executed based on the user’s prompt. This allows for more efficient management of the LLM context window and overcomes the MCP tool registration limits in Cursor IDE and Claude desktop app.
Custom MCP SSE added: You can now register external MCPs as Custom MCPs in AI Hub, enabling even more diverse MCP integrations.
Notion MCP added: Access and edit pages stored in Notion directly through the Notion MCP.

Improvements

Fixed QueryPie MCP integration errors
Fixed user UI page errors