text2future/flowix

What it solves

Flowix addresses the problem of fragmented context when working with AI agents. Instead of starting conversations from scratch or manually pasting information, it provides a centralized Markdown-based memory system where notes serve as durable, reusable context that agents can read from and write back to.

How it works

Flowix is a Markdown notebook application that stores notes as plain files locally on the user's device. It integrates with AI agents through a CLI and an MCP (Model Context Protocol) server, allowing both internal agents and external tools (like Claude Code or Codex) to search, read, create, and edit the notes. Users can explicitly control which notes or folders are shared as context for a specific task.

Who it’s for

It is designed for individuals managing complex projects—such as software developers, product managers, and researchers—who need to maintain a persistent knowledge base that AI agents can interact with to maintain continuity across different sessions.

Highlights

• Local-first storage: Notes are kept as plain Markdown files, ensuring user control and compatibility with other apps.
• MCP Integration: Supports the Model Context Protocol to connect various external agents to the same local memory.
• Bidirectional interaction: Agents can not only consume notes as context but also save their results back into the notes for future use.
• Granular access control: Users choose exactly which notes or folders are exposed to the agent.