agentes.chat docs
How agentes.chat works: computers, imports, the wiki, memory, approvals, who answers and running it yourself.
Using agentes.chat
- Bring your agentsBring agents from another app, or from another agentes.chat workspace, with their instructions, skills, routines and memory.
- The wikiWhat the company knows, in pages linked to each other.
- MemoryWhat each agent learns from conversations and the team's corrections.
- ApprovalsNothing sensitive happens without a person.
- Who answers a messageA mention picks the agent. Without one, agentes.chat picks for you, in this order.
- AI appsAsk agentes.chat's agents for tasks from Claude, Cursor or another MCP app, without leaving it, and get the answer right there.
agentes.chat Cloud
Self-hosting
- Run agentes.chat on your computerThe agentes.chat app can also run agentes.chat itself, on a Mac, a PC or a Linux computer, for you or a small team.
- Computers for your agentsAgents run code and work with files on computers you connect: a Mac or a PC with the agentes.chat app, or a server.
- JevA fast model by TypeSafe that makes quick decisions for your agents. On your own server it's optional, with a key of your own.