Agenticboard Your agents do the work.
Your board keeps up.
A Kanban board with docs built in — and an MCP server, so Claude, Claude Code and Cursor pick up tickets, follow your workflow, attach the evidence and move the card themselves. You keep the overview.
Works with Claude, Claude Code, Cursor, Claude Desktop, Sentry
Fix avatar overflow on the board
- Reproduced on the board view
- Fixed the overflow in TaskCard
- Tests pass — 1541 green
- Screenshot attached as evidence
Three steps to an agent that ships
Agenticboard speaks MCP, the open protocol AI clients use to reach your tools. No plugin to install.
Create a scoped token
In Settings → API tokens, pick read-only or read & write and which projects it may touch.
Point your client at it
One line for Claude Code:
claude mcp add --transport http \
agenticboard \
https://agenticboard.app/api/mcp/ \
--header "Authorization: Bearer agb_…"Hand it a ticket
The agent reads the board, follows the project's workflow, and moves the card when it's genuinely done. You review what changed.
Setup guides: Claude Code · Cursor · Claude Desktop · claude.ai · Sentry
Agents that follow your process
Give a project a workflow — reproduce, fix, test, attach evidence — and connected agents work the steps in order. They check what was already solved before re-solving it.
Boards and docs, one place
Kanban columns your team drags cards across, with Confluence-style nested pages right beside them. Agents read the same specs your people do.
Scoped and accountable
Tokens narrow to the projects you pick, read-only or read-write. Every call an agent makes is logged, and each card shows which agent filed it.
Simple pricing
Start free, on your own. Bring the team when you're ready.