Officiel MCP client. #4

louis2038 commented 2026年06月11日 10:26:38 +02:00

I know the repository has only recently been published and that the current priority is probably code stabilization, but I would still like to suggest a long-term feature idea.

It would be great to have an official MCP integration, something clean and well maintained, similar to the Logseq MCP project:
https://github.com/ergut/mcp-logseq

For me, this would be a real game changer. An LLM assistant becomes far more useful when it can access and navigate a structured personal knowledge base. With an MCP integration, the model could keep contextual awareness across different sessions, explore notes through wikilinks, and even generate intermediate reports during ongoing research work.

One of the biggest strengths of this kind of workflow is that the memory becomes explicit and inspectable. Users can organize and curate their own knowledge while the assistant traverses the graph structure naturally. It also reduces manual context injection and token consumption significantly.

Thanks for your work :)

I know the repository has only recently been published and that the current priority is probably code stabilization, but I would still like to suggest a long-term feature idea. It would be great to have an official MCP integration, something clean and well maintained, similar to the Logseq MCP project: https://github.com/ergut/mcp-logseq For me, this would be a real game changer. An LLM assistant becomes far more useful when it can access and navigate a structured personal knowledge base. With an MCP integration, the model could keep contextual awareness across different sessions, explore notes through wikilinks, and even generate intermediate reports during ongoing research work. One of the biggest strengths of this kind of workflow is that the memory becomes explicit and inspectable. Users can organize and curate their own knowledge while the assistant traverses the graph structure naturally. It also reduces manual context injection and token consumption significantly. Thanks for your work :)

joch commented 2026年06月11日 19:27:24 +02:00

@louis2038 Thank you for sending this suggestion and the Logseq example. I think you're probably right that this would be useful to people. I'll add this to the roadmap to explore. Not sure how soon I'll be able to look into it but it seems like something worth thinking about.

@louis2038 Thank you for sending this suggestion and the Logseq example. I think you're probably right that this would be useful to people. I'll add this to the roadmap to explore. Not sure how soon I'll be able to look into it but it seems like something worth thinking about.