Blog
Making context travel.
Practical guides on moving your working state between AI agents, connecting the tools you already use, and building local-first MCP servers.
Export your ChatGPT memory and give it to every AI agent
Every chat assistant you use builds up a picture of you: your rules, your preferences, your projects. None of them share it. Here is how to export that memory with one prompt, review it, and hand it to Claude Code, Cursor, Codex and every other agent you use.
Read →AI playbooks: turn your best agent sessions into reusable prompts
Every good session with an AI agent leaves a method behind, and almost all of it evaporates when the tab closes. A playbook is that method, captured as a reusable prompt. Here is why they matter now, how ctxfile distills them from your own sessions, and how the idea scales from one developer to a whole company.
Read →How to give Claude Code memory across sessions
Every new Claude Code session starts from the files on disk and whatever you wrote in CLAUDE.md. The plan, the decisions, and where you stopped are gone. Here is what actually persists, what does not, and a local fix that works with Cursor and Codex too.
Read →How to Share Context Between Claude Code and Cursor
You explain the project to Claude Code in the terminal, then open Cursor and explain it again. Neither tool can read the other's memory. Here is why, what the common workarounds get wrong, and how to give both of them one shared context layer.
Read →Claude Code /compact vs /clear, and How to Stop Losing Context Between Sessions
Both commands manage the context window and they are not interchangeable. Here is what each one actually does, when to reach for which, and the gap neither of them closes.
Read →How to Switch from Cursor to Claude Code Without Losing Your Context
The hard part of the migration is not learning the new tool. It is that everything Cursor knows about your project stays in Cursor. Here is what does not transfer, and how to carry it across.
Read →Connect Your Obsidian Vault to Claude Code, Cursor, and Any AI Agent
Your vault holds the reasoning behind your project. Your coding agent cannot see any of it. Here is how to give agents read-only access to the notes that matter right now, without uploading your vault anywhere.
Read →How to Build a Local-First MCP Server in TypeScript
Most MCP tutorials build a server that calls a remote API. This one builds a server that never leaves the machine, and covers the design rules that keep it that way.
Read →