The ultimate CLI agent management tool. Organize Claude Code, Codex CLI, and Gemini CLI from a single dashboard. Rename agents, switch editors instantly, and streamline your AI-powered development workflow.
Keep all possibilities open Don't trust AI guessing. Make it verify with actual server, DB, API calls
Read the functions your function calls Don't guess, actually trace the logic. Timing conflicts between functions cause one to get skipped
Search everywhere that function is used It's being called somewhere you didn't expect
Fixed a screen freeze bug that wouldn't go away for days using this method. The culprit was insufficient function call intervals + fit function not executing.https://www.solhun.com/changelog
What if you could manage all your projects and CLI agents in one place?
I got this idea while looking at Antigravity's agent manager. And watching all these AI tools constantly updating, I became convinced that I shouldn't be locked into any single AI or tool.
Amazing, finally someone did this. I was looking for something like this for a while. How does it manage restarts? Let's say I run some "export" commands, then start long running process. Then after restart, if I open CLI Manager again - what will I see? will it restart my processes? Will it preserve "env"? Will "history" be shared between tabs?
@avloss Thanks for the question! History is shared between tabs. Think of it as managing multiple terminals in one space. Env variables and long-running processes aren't persisted yet, but it's something I'm considering for future updates!
I picked this up today and am obsessed!! Feature request already. I want to have multiple terminals open at once and be able to have them split viewed inside CLI Manager.
You got me for life! Amazing work!
Report
Genius solution for managing multiple AI agents! As someone building AI-powered ITSM automation, this is exactly what we need - unified control over Claude, Codex, and Gemini agents. The instant editor switching is a game-changer for development velocity. Perfect fit for enterprise teams!
One thought from my experience building mobile testing frameworks at scale: the core problem often isn’t how many CLIs I manage, but whether any given CLI can reliably complete my actual task.
If a CLI (Codex, Claude, Gemini, etc.) can understand my project docs, task specs, or reusable “skills” and execute against them, then the CLI itself becomes interchangeable. In that world, the user’s task descriptions, project context, and skills become the real long-term asset — more valuable than the manager layer around the tools.
Curious how you’re thinking about evolving beyond CLI orchestration toward task- or skill-centric workflows over time.
CLI Manager
DeepTagger
Amazing, finally someone did this. I was looking for something like this for a while. How does it manage restarts? Let's say I run some "export" commands, then start long running process. Then after restart, if I open CLI Manager again - what will I see? will it restart my processes? Will it preserve "env"? Will "history" be shared between tabs?
CLI Manager
@avloss Thanks for the question! History is shared between tabs. Think of it as managing multiple terminals in one space. Env variables and long-running processes aren't persisted yet, but it's something I'm considering for future updates!
Any free trial? @solhun
CLI Manager
@lajlev Great question! Yes, we do have a free tier available!
Any plans for @Zed editor support?
Genius solution for managing multiple AI agents! As someone building AI-powered ITSM automation, this is exactly what we need - unified control over Claude, Codex, and Gemini agents. The instant editor switching is a game-changer for development velocity. Perfect fit for enterprise teams!
CLI Manager
@imraju Really appreciate the kind words! I'll keep updating it with features that actually help in real workflows.
Dessix
Interesting take on managing multiple AI CLIs 👍
One thought from my experience building mobile testing frameworks at scale: the core problem often isn’t how many CLIs I manage, but whether any given CLI can reliably complete my actual task.
If a CLI (Codex, Claude, Gemini, etc.) can understand my project docs, task specs, or reusable “skills” and execute against them, then the CLI itself becomes interchangeable. In that world, the user’s task descriptions, project context, and skills become the real long-term asset — more valuable than the manager layer around the tools.
Curious how you’re thinking about evolving beyond CLI orchestration toward task- or skill-centric workflows over time.