LightfieldAI-native CRM that builds itself and does work for you
Promoted
Masko Code
A mascot that watches Claude Code for you
274 followers
A mascot that watches Claude Code for you
274 followers
You run Claude Code agents. You alt-tab 50 times a day to approve permissions. You lose your place. You miss prompts. Masko Code puts a mascot on your desktop that watches your agents in real time. Permission needed? Speech bubble. Press ⌘1. Done. Also: ⌘M jumps to the right terminal. Double-tap ⌘ to switch sessions. Tracks everything at a glance. Ships with Clippy and other mascots. Community adds new ones daily. Or generate your own at masko.ai. Free. Open source. MIT. 100% local.
Love that this started as a fun mascot idea and evolved into an actual productivity tool. The ⌘1 to approve permissions without leaving the editor is the real feature here. Any plans to support other AI coding agents beyond Claude Code?
@mehmet_kerem_mutlu Yes codex will come very soon
@hex_miller_bakewell Yes! It hooks into Claude Code through its native hook system. The mascot runs on a state machine graph so each state (idle, working, attention, thinking) can trigger different animations and actions. The goal is to make everything fully customizable. With masko ai you can generate any character in any style, and the interactions are endless because you control the states. Community adapters for other tools is definitely where this is heading. Love the multiple mascots idea!
I keep running into this and can't figure out a good way to handle it. Claude shows "thinking" but it's unclear whether it's actually working or stuck in a loop, maybe hitting some architecture issue underneath. Is there any detection for that in Masko? Like, can it tell the difference between "model is processing" and "model is retrying silently and going nowhere"?
@slavaakulov today it tracks the agent state through claude code hooks so you can see when Claude is thinking, working, or spawning subagents. There's a full trace of all tool uses in the notification feed. But it can't tell i think the difference between "thinking normally" and "stuck in a silent retry loop." That distinction would need deeper hook support from Claude Code itself.
Sounds like a great product! Running multiple Claude Code sessions can be very chaotic. This solves the exact thing that breaks the flow most times. Any plans to support any other AI coding agents beyond Claude Code?
@simonk123 Thanks Simon! Yes, the architecture is agent-agnostic. Right now it hooks into Claude Code but supporting Codex, Gemini CLI and others is on the roadmap.
This is one of those 'why didn't this exist already' tools. The alt-tab tax with Claude Code agents is real and nobody was talking about it. Love that it's fully local with zero telemetry that's the right call for a developer tool. Two things I'm curious about: any plans to support watching multiple agent sessions at once? And is there a way to customize what triggers the speech bubble (like only for permission requests vs. all pauses)? Also the fox is adorable, not gonna lie.
@samet_sezer To answer both: yes, it already tracks multiple sessions simultaneously. Double-tap ⌘ to see all active agents and switch between them.
The speech bubble currently triggers on permission requests and questions, but customizing trigger types is a great idea. And thanks, the fox appreciates it.
The speech bubble for permission requests alone would save me a lot of broken focus. Love that it's open source and fully local too. Any plans to support other coding agents beyond Claude Code, like Cursor or Copilot terminals?
@ben_gend Thanks! Yes, Codex support is coming next with help from the community. For Cursor and Copilot it depends on how easy it is to hook into their event systems, but we're looking into it.