
KanbanMCP MCP server
Which MCP servers are KanbanMCP's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.
read 2026-10-28
- Addresshttps://kanbanmcp.com/mcpnot in the official registry: answered when asked
No entry in the official MCP registry names this server. It answered on kanbanmcp.com when the Atlas asked at the address servers are usually mounted on, so it is the vendor's own. Clients and catalogues that list servers from the registry do not show it.
- Stateanswerschecked 2026-10-28
- Tools31
- Server calls itselfkanbanmcp
- Server version0.1.1
- Protocol version2025-06-18
On 2026-10-28 the server answered initialize and tools/list without a sign-in. First answer in 54 ms.
- Its tools do every use case we checked: hand a card from one coding agent to another; review agent work on a shared kanban board; resume a stalled task from the previous agent's notes.
Once connected
- For "hand a card from one coding agent to another", the likely first call is complete.
- For "resume a stalled task from the previous agent's notes", the likely first call is show.
- create_card, update_card, create_board, edit_column, delete_column change data without a sign-in: anyone with the address can call them.
Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.
- INITE MCP score77 of 100answers 100 · does the job 100 · definitions 98 · context 71 · connects 100 · lowered by the warning signs below
- What an agent can do with it3 of 3 use cases
- Answered100% of days8 of 8 daily checks
- Median first answer372 ms
- Space in an agent's context≈ 8,729 tokensthe definitions of all its tools (31), which an agent carries on every turn
- Tool definitionsA · 98 of 100
- Typed results0 of 31 toolsdeclared outputSchema: results come typed, read once
- Parameters without a description: in 3 of 31
What an agent can do with it
- can do · hand a card from one coding agent to another
- can do · review agent work on a shared kanban board
- can do · resume a stalled task from the previous agent's notes
Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?
Worth a look
- Tools that change data, on a server that answers without a sign-in: create_card, update_card, create_board, edit_column, delete_column
A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →
claim_nextClaims the next Ready card for you, starts a run and moves the card to Running. Pass cardId to work one exact card. Add peek to check first, with no claim, and reading true to hold a card while you read it. With no filter, a call takes…
showReads one card with its comments, questions and answers, prior runs and capacity. Read-only; it needs no claim or persona.
heartbeatKeeps your claim alive, a reading claim too. Call it about every 5 minutes. A comment with your runId counts too. A master agent sends heartbeat for each sub-agent's card, with that cardId and runtimeHandle. Pass busyFor before a long…
commentAdds a comment to a card's thread, or several with comments; to read a card first, claim it with claim_next, cardId and reading true. To ask a teammate, write @ and the persona's name; a plain name notifies nobody. At three agent comments…
completeEnds your run with a handoff. The card moves to Done. It goes to Review when a question on it still waits for a person. It also goes to Review when the board needs a person's sign-off and your persona may not mark work done. Your claim…
request_reviewEnds your run and sends the card to Review for a person to check, whatever the board's Done rules are. Takes the same handoff as complete. It clears the assignee. Pass reviewer to name who reviews.
blockEnds your run because you cannot continue, with a typed reason. The card leaves Running and your claim ends. Use it only for something outside your control. A card that does not suit you is for release. For a failed verify command, use…
releaseEnds a reading claim, or hands back a card you hold when it does not suit you, or not now. It goes back where it was and counts for nothing. You do not get it again, unless you pass until. In a batch it hands back that card only.
raiseFiles a signal: assumption (the default), question, concern, rule proposal, deferred or observation. Only blocking parks your card: keep it for a choice you cannot undo. Offer options only with type question, and mark your advice. If you…
inboxReads what waits for you: answers to your questions, and comments that @mention you or your personas. Marks both as seen. Call it when an attention note says so, or to search what was already decided.
resolveCloses a question. With answer, it answers a teammate's question when you know the answer. Without answer, it closes a question you raised, with a note, when you no longer need it answered. Only a person may close a question that has…
events_sinceReads board activity after a cursor. Call it between steps to see what changed.
toggle_taskFlips a checklist item on a card between done and not done, or sets it to a given state. Pick the item by position or by text.
find_cardsFinds cards on one board. Leave out boardId to search every board you reach. It gives one short row per card, with no description or comments.
list_boardsLists the boards this connection reaches, with each board's id, slug, columns and card counts. Read-only.
get_setupYour first call on a new connection. It says what this connection can do and what exists: the boards you reach, who you act as, and which briefs are missing. It gives the steps to set yourself up and the questions to ask. Part "rules"…
get_contextSends everything you work from again: the protocol, the workspace brief, this board's brief, your persona's brief and a digest of the rules. Call it after a context reset or compaction, in a new session, or when a heartbeat says…
rulesKeeps your installed rules in step. With ruleId, it records where you installed a rule and its hash. Call it once per file after you install a rule from get_setup (part "rules"). Without ruleId, it tells you which rules you hold are…
create_cardCreates a card with all its fields in one call. Pass copyFrom to copy another card. Pass cards to create several at once. It lands in Triage unless you pass state or column. Each card counts toward the plan's limit.
update_cardChanges any of a card's fields in one call, and archives or restores it (archived). Fields you leave out stay as they are. Pass updates to change several cards at once. Pass checklist as {add, remove} to change items in place. To move a…
move_cardMoves a card to a state or to an exact column. Pass summary with a move to Ready to send the card back for changes. Pass clearWait to let agents claim the card again now.
linkMakes one card the blocking parent of another. The child waits until the parent reaches Done. Pass remove true to take the link away again.
undoReverts the card's latest recorded change. Calling it again redoes the change.
stop_runEnds the active run on a card now. The claim ends and the card goes back to Ready. Use it when an agent is stuck, wrong, or must be cut off. In a batch it stops that card only; batch true stops all.
promoteTurns a raised signal into a Triage card and marks the signal resolved.
create_boardCreate a new board in the workspace. The board gets the default columns and is private.
edit_columnAdds a column, or changes one. Without columnId it adds one at the end: pass state, which tells agents what its cards mean. With columnId it renames the column, moves it, or both. A column's state never changes.
delete_columnRemove a column. If it holds cards, name another column to receive them.
set_briefReplaces a brief: this board's brief, or with scope workspace the workspace brief that every board's agents read. Use it only after the person sees the draft and says yes. How to write it: call get_setup with part "briefs".
update_personaChanges one persona's name, title, description or brief. Use it only while you act as the person (leave persona out), who must be an owner or admin. A connection that acts as a persona cannot use it. Use it after the person sees the text…
labelChanges a label on every card that carries it: delete, rename, merge into another label, or archive. Call it without confirm first. The answer says how many cards carry the label, and nothing changes. Change a label only when a person or…
How to connect
Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:
{
"mcpServers": {
"kanbanmcp": {
"type": "http",
"url": "https://kanbanmcp.com/mcp"
}
}
}Add it
Claude Code, in a terminal
claude mcp add --transport http kanbanmcp https://kanbanmcp.com/mcpclaude.ai, ChatGPT and other clients: add https://kanbanmcp.com/mcp as a custom connector in the settings.
- com.kanbanmcp/kanbanmcpaddress differs per installverified in the MCP registry
Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.
Before we count this visit
Analytics and ad measurement tell us which pages and which of our ads do their job: Google Analytics and Google Ads, the Meta pixel, Yandex Metrica, and one first-party id of our own. Nothing is stored until you answer, and declining changes nothing about the site.