The latest articles on DEV Community by 南羽 (@nanyujun). https://dev.to/nanyujun https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4221487%2Fd46541ad-5697-4a3d-8939-8e77ce9c8deb.jpg https://dev.to/nanyujun en 南羽 Mon, 02 Nov 2026 04:09:24 +0000 https://dev.to/nanyujun/keeping-ai-coding-agent-sessions-in-sync-across-machines-5h7 https://dev.to/nanyujun/keeping-ai-coding-agent-sessions-in-sync-across-machines-5h7 <p>I work across more than one machine, and I switch between coding agents depending on the task: Claude Code, Codex CLI, Cursor, Gemini CLI. The models keep getting better, but the context is still nailed to one laptop.</p> <p>Every agent stores its history differently, and locally:</p> <ul> <li>Claude Code keeps CLI transcripts under <code>~/.claude/projects/&lt;encoded-project-path&gt;/&lt;session-id&gt;.jsonl</code> </li> <li>Other agents have their own session directories and formats</li> <li>Cursor keeps chat state inside its workspace storage</li> </ul> <p>Copying the project folder is not enough, because session files live outside it. Copy them to a new machine and the paths may not match either. So "continue where I left off" turns into manual file archaeology.</p> <p>I built AIAssets to fix that for my own workflow.</p> <h2> What it does </h2> <p>AIAssets is a self-hosted hub. On each machine it scans the session files of the agents it supports, indexes them, and syncs the index and session data bidirectionally through the hub on a 60-second poll. Session titles are preserved, so you can find the thread you were in. Over that index it also gives you search, share, and backup.</p> <p>Self-hosting is a deliberate choice: transcripts contain source code, internal URLs, and sometimes commands with credentials in them. I did not want a third-party service holding all of that.</p> <h2> Support and current limits </h2> <p>The README lists 10 agents with read/write support, including Claude Code, Codex, Cursor, and Gemini CLI. It is early software, so the limits matter more than the feature list:</p> <ul> <li>v1.0.0 ships a Windows installer only.</li> <li>On macOS you build the DMG from source with <code>pnpm run build:mac</code>; it is not signed, so Gatekeeper will complain.</li> <li>Sync is polling, not real-time.</li> <li>There is no conflict UI yet. If you continue the same session on two machines, I do not have a good merge story.</li> <li>Agent updates change session formats, and parsers break.</li> </ul> <h2> The parts I am unsure about </h2> <p><strong>1. Project path mapping.</strong> A session is tied to a working directory. If machine A uses <code>/Users/me/code/foo</code> and machine B uses <code>D:\code\foo</code>, restoring on B can show the session but fail to resume because the cwd does not exist. Should the hub rewrite paths, or store a logical project ID? I lean toward the latter, but it is not shipped.</p> <p><strong>2. Divergent branches.</strong> The same session continued on two machines creates two files. Last-write-wins loses one side. A real merge of JSONL transcripts is harder than it sounds, and I would rather not fake it.</p> <p><strong>3. Size and transport.</strong> Sessions can get large. Polling the full hub every 60 seconds is fine on a LAN and wasteful over a metered connection. Incremental transfer is on the list.</p> <h2> Repo </h2> <p><a href="https://github.com/NanYuJun/AIAssets" rel="noopener noreferrer">https://github.com/NanYuJun/AIAssets</a></p> <p>If you run more than one machine or more than one agent, I would rather hear which part breaks first than get a star. Issues with your setup — OS, agents, where the sessions live — are more useful than "doesn't work".</p> ai claude opensource productivity