Guides and reference for the app at app.kadmo.ai, its agents, workflows, integrations, MCP and the Kadmo provider, and for the runtime on your own computer.
Kadmo runs AI coding agents for a software team. You describe the work, or hand Kadmo a ticket; an agent with the right role picks it up, follows the playbook for that kind of task, and stops at a gate whenever a person should look. Every step leaves its record: the run's transcript in the app, and a comment on the ticket.
| Part | What it does | Read more |
|---|---|---|
| Account | Members, roles, settings and billing | Account and settings |
| Workspaces | Where work happens: path, repos, instructions, routing | Workspaces |
| Agents | The machines that run the work | Agents |
| Ask Kadmo | Answers from your account and turns a request for work into a plan | Ask Kadmo |
| Tasks and playbooks | Tickets in, dependency-ordered runs out | Working with tasks |
| Automations | Ask Kadmo where you work — a @kadmo comment or a label in Jira, a message in Slack or Teams — and rules that start work from tracker and git events | Automations |
| Reporting | What agents and Kadmo write back on the ticket, in Slack, in Teams and by e-mail, and how to silence it | Reporting |
| The Kadmo provider | Models, credit and budgets | Models and credit |
| MCP | Use your account from an AI client | MCP overview |
| The Kadmo runtime | Run Kadmo on your own computer: the kadmo command, its local MCP server and the Kadmo extension for Chrome | The Kadmo runtime |
