The shift from documentation as a passive reference to documentation as active context for agents is one of the biggest changes in knowledge management right now.
If your notes aren't queryable by an LLM, they're becoming a graveyard... The old adage "people don't read the docs" which was never really true, might now actually become reality. (Yes, I did work as a documentation manager at one point in my career, does it show?)
So for my own sideproject, I built Vist with an integrated MCP server for exactly this reason: it turns your project notes into a structured memory for Claude, or Gemini, or ChatGPT. It’s less about 'writing for humans' and more about 'architecting context' for the agents we're now working alongside, present and future.
The shift from documentation as a passive reference to documentation as active context for agents is one of the biggest changes in knowledge management right now.
If your notes aren't queryable by an LLM, they're becoming a graveyard... The old adage "people don't read the docs" which was never really true, might now actually become reality. (Yes, I did work as a documentation manager at one point in my career, does it show?)
So for my own sideproject, I built Vist with an integrated MCP server for exactly this reason: it turns your project notes into a structured memory for Claude, or Gemini, or ChatGPT. It’s less about 'writing for humans' and more about 'architecting context' for the agents we're now working alongside, present and future.