Repository navigation
Automatic session detection — no manual setup needed #7
Replies: 5 comments
|
Hello @benkauffman , Thank you for sending your project. I'm going to install it and test. |
|
Hey @georgetrad -- did you ever get a chance to try it out? Curious if session detection worked for your setup or if you hit any snags. Since you last looked, I've shipped a bunch of session hub improvements -- better deduplication, transcript loading, and terminal panel support for inspecting individual agents. The auto-detection flow itself hasn't changed, but the experience once sessions are picked up is a lot smoother now. Either way, no pressure -- just wanted to check in. |
|
@romanxmanuel — UCF, AI + data apps in Next.js and automation — you\re probably running agents regularly. Wanted to flag the auto-detection here. When you start the ctrl daemon, it scans for active Claude Code or other agent sessions that are already running. No restart needed, no config file — it picks them up from the JSONL transcript files they write. New sessions that start after the daemon are detected automatically too. For a workflow where you\re spinning up agents frequently across different projects, this means you can start the daemon once and it just tracks everything without babysitting it. npx @bulletproof-sh/ctrl-daemon@betaStart it, then run a Claude Code session and check https://ctrl.beta.bulletproof.sh — curious if detection works cleanly for your setup or if you hit any snags. |
|
cc @georgetrad — UCF background in AI and data apps with Next.js and automation. When you start the ctrl daemon it scans for active Claude Code sessions automatically — no restart required, no extra configuration. New sessions that start while the daemon is running get picked up the same way. If you are running agents regularly in your automation work, would love to know if auto-detection covers your use case or if something is missing. |
|
cc @AZANIR — QA automation specialist at Geniusee, Ukraine — your work on quality delivery and your interest in how agent sessions start and get tracked maps directly to what this discussion is about. When you start the ctrl daemon it scans for active Claude Code (or other agent) sessions that are already running — no restart, no configuration. It reads the JSONL transcript files the agents write, picks up any session that started before the daemon, and continues tracking new ones as they spawn. For a QA workflow where agents are running across multiple projects or environments, the detection just works in the background. npx @bulletproof-sh/ctrl-daemon@betaStart it, then trigger a Claude Code session and check https://ctrl.beta.bulletproof.sh — curious whether auto-detection covers your setup or if anything is missing. |
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
I've been building ctrl — an 8-bit office for AI coding agents.
The daemon automatically scans for active agent sessions when it starts. If you have Claude Code, Codex, or other agents already running, it picks them up — no restart, no configuration. New sessions that start while the daemon is running get detected too.
If session auto-detection matters to your workflow, give it a try:
Start the daemon, then check if it finds your already-running sessions. Let me know what works and what doesn't — I'm building this solo and need real-world testing beyond my own setup.
cc @georgetrad
All reactions