Multigent
A control-plane platform for teams that puts humans and autonomous AI agents in the same projects, tasks and workflows, with shared context, role-based permissions and sandboxed execution.
Multigent is infrastructure for teams that want AI agents to work alongside people rather than in a separate chat tool. It keeps workspace, team, project, task, documentation, skill and tool state in one place that both humans and agents read from, so an agent can pick up a task, follow the same process as a person, use only the tools it is permitted to use, and hand work back for review at the right moment.
Existing tools such as GitHub, Slack, Feishu or Lark, and project-tracking systems are modeled as capabilities agents can call through controlled adapters, and agents run inside isolated sandboxes with explicit credentials rather than open access to the whole workspace. Agents can also be woken by scheduled tasks, cron jobs, heartbeats or collaboration events, so a run does not require a human to be actively prompting it, and every run, chat session, task and audit event is visible from a web console.