What is Board Planner
Board Planner is an issue tracker built around one idea: the board is an interface for people and for agents at the same time.
In most trackers, automation is bolted on — a bot posts comments, a script moves a card. Here the board, the REST API and the MCP server are three doors into the same model, with the same permission checks behind each of them. An agent that moves a task to In Review goes through the same status rules as a person dragging the card.
Who it is for
Section titled “Who it is for”- Small product teams who want a fast Kanban board without the ceremony of a large tool.
- Developers working with coding agents who want the agent to read the backlog, pick work up, and report back where the team can see it.
- Anyone automating around a board — the REST API and webhooks are first-class, not an afterthought.
What makes it different
Section titled “What makes it different”| Capability | What it means in practice |
|---|---|
| Columns with roles | You name your own columns, but each one is tagged backlog, approved, active, review, blocked or done. Automation keys on the role, so renaming a column never breaks a workflow. |
| MCP server | Claude Code and other MCP clients get twelve tools for reading and changing the board — no scraping, no glue code. |
| PM agent | An in-app project manager you chat with, which can also review the board on a schedule and act on what it finds. |
| Execution workers | A worker process claims approved tasks, runs a coding agent in an isolated checkout, and only pushes if the merge gates pass. |
| Human gates | Any column with the review role is a stop sign for automation. Work waits there for a person. |
| Scoped tokens | An API token can be limited to specific projects. The limit is enforced centrally, so it holds for REST and MCP alike. |
What it is not
Section titled “What it is not”- Not a document or wiki tool. Task descriptions take Markdown; long-form docs belong elsewhere.
- Not a time-tracking or billing product. There are no timesheets.
- Not multi-tenant SaaS. One deployment is one instance, with its own users and projects.