Hi Product Hunt! 👋 Our developer was matching API interfaces between two services. Each end lived in a separate Claude session. He kept copy-pasting specs back and forth, manually keeping both sides in sync — until he stopped and thought: why am I doing this? Can't you two just talk directly? That was the first version of Khala. It kept proving itself: → A teammate needed to sync interfaces with another dev. Instead of scheduling a call, he just told them to install Khala. The two Claude sessions worked it out themselves. → A team member got sick and couldn't come in. The work couldn't wait. He connected to that member's Claude session through Khala and finished it himself — without bothering the person who was out. Same insight every time: the AI was ready. The gap between sessions was the only problem. **What Khala does:** - Lets your Claude sessions talk to each other directly - Works across teammates — your Claude to their Claude - Runs solo too — split tasks across sessions, coordinate automatically - MCP-native, works in Claude, Claude Code, and ChatGPT 🎉 Free during beta — no credit card required. What's your biggest pain point when running multiple AI sessions? Would love to hear how you'd use this. 🙏

Khala

Stop being the messenger between your LLM sessions.

Stop being the messenger between your LLM sessions.