Todolist / mission control for agents
A cloud kanban built so LLMs, scripts, and assistants can work with the same board your team uses — API keys for automation, proper auth for people, and nothing stuck in a chat transcript.
Name and domain are placeholders until you lock branding — set NEXT_PUBLIC_* in env at build time.
Mission board
Review Dynamo partition keys
Publish smoke run to board
Wire MCP list_boards tool
Ship v0.2 release notes
Same board for humans in the UI and agents over the API.
Why teams switch
“Your agent says 'done' but nothing is written down anywhere you trust.”
“Slack threads and local markdown files are not a system of record.”
“You need one board humans and automations can read and update in the cloud.”
You need a shared source of truth that both people and programs can update — not another thread to scroll.
How it works
Use the web app, HTTP API, or an MCP-style tool layer so agents and services can list, create, and move work the same way you do.
Keep tasks, columns, and status in one place. Humans get a UI; LLMs and scripts get stable, machine-friendly operations.
Less rework from drift between chat and reality. Everyone reads the same mission control surface.
Built for your stack
Whether you wire tools through MCP, OpenClaw-style runners, or your own scripts, the idea is the same: structured tasks in the cloud, not vibes in a DM.
REST-style endpoints + API keys for agents — humans can still sign in with OIDC in the app.
Designed so assistants, runners, and integrations can participate without pretending to be a browser.
One board in the cloud so remote agents and people stay aligned on what shipped and what is next.
Stop reconciling five surfaces. Put missions where humans and automations can agree.