Feature Requests

Description: It would be great to add support for Anthropic's Model Context Protocol (MCP) to MindMac. MCP is an open standard that enables AI models to connect with external tools, data sources, and services through a unified protocol. Benefits: • Extended functionality — Users could connect MindMac to external tools (file systems, databases, APIs, web browsers, etc.) without leaving the app • Standardized integration — MCP provides a consistent way to add new capabilities without custom implementations for each tool • Enhanced productivity — AI assistants could perform real-world actions like reading/writing files, executing code, querying databases, or interacting with third-party services • Future-proof architecture — As MCP gains adoption across the AI ecosystem, MindMac would stay compatible with emerging tools and servers Suggested Implementation: • Support for MCP server configuration (local and remote) • Ability to enable/disable specific tools per conversation or globally • Tool approval workflow for sensitive operations • Display of tool calls and results in the chat interface Use Cases: • Connecting to local development environments • Integrating with note-taking apps and knowledge bases • Accessing real-time data from APIs • Automating workflows with external services This feature would significantly enhance MindMac's capabilities and align it with the growing MCP ecosystem already supported by Claude Desktop, Cursor, and other AI tools.

Description: It would be great to add support for Anthropic's Model Context Protocol (MCP) to MindMac. MCP is an open standard that enables AI models to connect with external tools, data sources, and services through a unified protocol. Benefits: • Extended functionality — Users could connect MindMac to external tools (file systems, databases, APIs, web browsers, etc.) without leaving the app • Standardized integration — MCP provides a consistent way to add new capabilities without custom implementations for each tool • Enhanced productivity — AI assistants could perform real-world actions like reading/writing files, executing code, querying databases, or interacting with third-party services • Future-proof architecture — As MCP gains adoption across the AI ecosystem, MindMac would stay compatible with emerging tools and servers Suggested Implementation: • Support for MCP server configuration (local and remote) • Ability to enable/disable specific tools per conversation or globally • Tool approval workflow for sensitive operations • Display of tool calls and results in the chat interface Use Cases: • Connecting to local development environments • Integrating with note-taking apps and knowledge bases • Accessing real-time data from APIs • Automating workflows with external services This feature would significantly enhance MindMac's capabilities and align it with the growing MCP ecosystem already supported by Claude Desktop, Cursor, and other AI tools.

The claude 4 models are really critical these days and need to be added to api support.

Someone post this request: https://mindmac.canny.io/feature-requests/p/refreshupdate-models-for-ollama-and-perplexity I think you should give update models icon for every endpoints. For example, recently we have GPT-4o running, but I have to add it manually. Why dont it has refresh model icon? Or how if there's only one button for "one click to refresh" all models from all providers? Thx.

Can perplexity, openrouter, together, ollama, etc. which are currently supported by the app in Global mode be made to support to inline mode too?

Currently we can use text to image, but we can't upload an image and regenerate it.

For users of multiple devices, it could be very convenient to have the same favorites, folders and conversations available on all devices. This would require a sync function, or saving the conversations, folder data and most of the settings in a project file. The latter option (project file) would only be an alternative idea if a sync function is too much work for the current state. Then the user himself can simply save the project file on his own cloud share/drive. Once a project file is defined, MindMac could check automatically for changes or update the project file when some are done. This is fore sure, a bigger feature, so no stress, just for road-map documentation. ;D Suggested by Sascha

I'm wondering if the app development is still ongoing, as I have not received any replies to my support requests.

I really like your software, and thank you for providing such an excellent tool. I’m currently facing a minor issue and would like to consult with you to see if it can be resolved. I’m using the OpenRouter API with the model deepseek:r1, corresponding to the model link: https://openrouter.ai/deepseek/deepseek-r1 . In OpenRouter’s official chat room, I can see the reasoning process of deepseek, but when using MindMac, I could not see the reasoning process. I would like to know whether MindMac supports displaying the reasoning process?