Everyone on your team uses AI differently: different tools, habits and subscriptions. ahel is the shared setup under all of them. Connect your team's apps and pick its skills once, and each person's AI picks them up, in Claude, ChatGPT, Cursor and more.
Replies
Best
Maker
📌
Hi Product Hunt. I am Karl, and I build ahel from Tallinn.
AI made me fast on my own. It did not make my team fast together. Everyone uses it differently: some of us ship as quickly as we can, some review every line and keep the docs right. Some live in Claude, some in ChatGPT, some in Cursor. So the way we work never landed in anyone's AI. We wrote handover docs, pasted the same instructions into every tool, set up the same connections three times, and still nobody's AI knew what we decided last week.
ahel is the shared setup under all of that. Your team connects its apps and chooses its skills once, in one workspace. Each person's AI reads them, whichever of the six supported AIs it is. Add an app or a skill, and it is there for everyone the next time they ask. Nobody switches tools and nobody stops working.
Plain facts: a connection one member makes is usable by the whole workspace, so connect work accounts, not personal ones. You cannot write your own private skills or upload your own documents yet. If you run your own server, you can connect that.
What works today: the apps marked Verified on ahel.ai/apps, which my team and I use daily. The rest are built from the vendors' own documentation and not yet tried by hand, and each page says so.
For the technical readers: it speaks MCP under the hood, the open standard AI tools use to talk to other software. You do not need to touch it.
One ask: tell me the last time your team's AI setups fell out of sync, and what you did about it. That is the list I build from.
Replies