Todokan is the kanban board where humans and AI agents work together — context on the card, one click to hand off, your approval on every step.
This listing identifies a third-party MCP server. MCPBundles is not affiliated with, sponsored by, or endorsed by the service or its operator unless stated on this page.
This server may ask you to sign in when your AI client connects.
Last probed Nov 3, 2026 · todokan.com
105tools discovered
For AI agents — MCPBundles Directory API
For MCP client apps (human connect):
MCP connect URL: https://todokan.com/mcp
Skill page: https://www.mcpbundles.com/skills/todokan-mcp-34ebace1f3
Explore capabilities (105)
read 13
write 12
Showing 25 of 105
Showing 25 of 105 tools from the public preview. Connect the server to inspect the full live tool list.
List6
Create3
Get3
Search2
Other capabilities11
List Habitats
read
list_habitats
Use when you need the user's habitats. Habitats are top-level containers that group related boards. Returns habitats as {id, name, version}. Read-only.
List Habitats — Use when you need the user's habitats. Habitats are top-level containers that group related boards. Returns habitats as {id, name, version}. Read-only.
Create Habitat — Create a new habitat in the user's Todokan account. A habitat is a top-level container for organizing related boards (Thought boards and Task boards). Name ≤80 chars; idempotent (returns existing on duplicate).
List Boards — Use when you need the user's boards or a boardId for another tool. Returns boards as {id, habitatId, name, version, agentEnabled, aiAccess, canCreateTasks, boardType, isSystem, deletable}. Route a task to the right board by matching its name (e.g. "Coding"). Before writing, check `canCreateTasks`: a board with aiAccess 'read_only' rejects create_task, and rejects having a task moved onto it too — do not try to work around it. Read-only; optionally scoped to habitatId.
Create Board — Create a new TASK board in a habitat (e.g. "Coding", "Marketing") to organize tasks. Requires habitatId and a name unique within the habitat. Planner only. Name ≤100; needs Tasks module; idempotent.
Update Board — Rename a user-created board. System boards (Tasks/Thoughts) cannot be renamed. Planner only. Name ≤100; rename only.
Delete Board — Delete a user-created board. System boards (Tasks/Thoughts) cannot be deleted. A board with tasks requires force=true (this also deletes its tasks). Planner only. Pass force:true to delete a board that still has tasks.
List Tasks — Use when the user asks to list tasks or you need task IDs. Optional filters: boardId, status, label(s), assignee. Use limit to cap results; cursor is only supported when boardId is set. Read-only; paginated (default 200, max 500); filter by status/assignee/labels.
Get Cost Summary — Read AI + external API cost totals over a period. Omit habitatId for the all-habitats portfolio total (full-access keys only; a habitat-scoped key is coerced to its own habitat). Returns aiCostCents, externalCostCents, totalCostCents and breakdowns byAgentRole / byProvider / byDay (plus byHabitat in all-habitats mode). Read-only; window 1–90 days (default 30).
Get your MCP into directories
A working endpoint is step one. Directory coverage is the coordinated launch across ChatGPT, Claude, Cursor, the MCP Registry, and community indexes.
Todokan is the kanban board where humans and AI agents work together — context on the card, one click to hand off, your approval on every step.
How do I connect Todokan to my AI agent?
Use the MCP endpoint listed on this page in your MCP client configuration. One-click install pills support Claude, Cursor, VS Code, and other hosts. Copy the remote MCP URL if your client needs a manual entry.
How many tools does Todokan provide?
Operate this server?
Request removal if you do not want Todokan listed here, or claim the listing if you want to manage it on MCPBundles.
Operate Todokan?
This server appears in the MCPBundles directory. Verify you operate it to take over the listing — name, description, logo, contact email, and skill content. We email a 6-digit code to a maintainer address your server publishes in /.well-known/security.txt or /.well-known/mcpbundles.json. Free, takes about a minute.
MCPBundles publishes this directory for MCP discovery. Except where we host or operate an offering, third-party services run under their own terms. Product and company names on this page are used in a descriptive, identifying way (including under nominative fair use where applicable); they remain the property of their owners. Nothing here grants you rights in those marks, and nothing here is an offer to sell a third party's services. Terms · Request removal
List Cost By Task — List the tasks that incurred the most AI cost over a period (cost, run count, last run). Omit habitatId for all habitats. Each task links to its session via /debug/agent/<taskId>. Read-only; window 1–90 days; limit max 100 (default 20).
List Cost Events — Chronological feed of cost events (each AI agent run and each external spend), newest first. Omit habitatId for all habitats. Use to see how costs are developing over time. Read-only; window 1–90 days; limit max 500 (default 100).
Search Context — Semantic search over the habitat's knowledge corpus (uploaded documents). Returns the most relevant text chunks for a query, scoped by your grant (RLS-enforced — you only ever see what you may see). Use to pull background/context for a task. The habitatId argument can only NARROW within your existing access, never widen it. Read-only; query ≤400, ≤12 chunks; needs knowledge module; RLS-enforced.
Search Across Habitats — Search tasks across all habitats/boards with one call. Use when the user asks for a topic across multiple boards (e.g. investor meeting, roadmap, hiring). Query ≤200, limit ≤100; tasks only; respects protection levels.
Get Events Since — Return a unified event feed since a timestamp (task lifecycle events, task comments, and documents). Use for polling-based agent monitoring. Since = ISO-8601 w/ offset; limit ≤500.
List Board Labels — List labels used by tasks in a specific board. Returns unique labels and usage counts. Read-only.
Propose Task Variants — Use when the user wants to add a new task. Generates 2–3 suggested task drafts (short/standard/detailed) for the user to pick from before creating. Title ≤160, description ≤500; returns variants to confirm before create.
Confirm Task Fields — Use after propose_task_variants. Shows a confirmation form the user can edit before task creation. Preview step before create_task; variantRef can go stale.
Get Task — Fetch a single task by ID. Returns the task as {id, boardId, title, description, status, assignee, aiEnabled, priority, labels, dueDate, version}. Use this when you have a taskId and need the full task record (for example before update_task or to read the brief). Returns -32003 if the task is missing, out of scope, or protected. Read-only.
Create Task — Create a task in the user's Todokan account. Use only when the user explicitly asks to create/add a task. If the board is missing, call list_boards and ask which board to use. Labels can include colors as hex codes (e.g. { name: 'urgent', color: '#ef4444' }). Title ≤80, description ≤500, ≤10 labels; needs boardId/boardName; idempotent.
Add Mission Context — Copy an EXISTING card into a mission as a read-only CONTEXT card. The copy lands in the mission board's 'Kontext' column (status 'closed', label 'mission-context') and records its origin (provenance appears in the mission's concentrate). Use to give a mission's agents the relevant cards as context. Cards from protected habitats are refused for agent callers. missionId = the mission habitat id; sourceTaskId = the card to copy in (may live in another habitat).
Submit Mission Context — Context-Scout submit: when dispatched in MISSION_MODE=scout inside a mission's MOTHER habitat, submit the source ids worth copying into the mission as read-only context. The server materializes them (cards → context cards; documents → shared-blob copies; emails/calendar events → context documents). Only source types the launch enabled are accepted; everything must live in the parent habitat. Call exactly ONCE per scout run.
Launch Mission — Launch a MISSION: a temporary child habitat where an agent team works a goal on a watchable kanban board. Creates the workspace + root goal card, seeds optional context cards, sets an optional cost ceiling (appetiteUsdCents), and dispatches the orchestrator to prepare a plan. The mission then waits in 'ready' for the human to review the plan and press Start. Results come back as a concentrate + proposals when the mission closes.
Start Mission — Start a prepared mission (status 'ready' → 'running'): approves the plan on the board and dispatches the orchestrator to staff the subtasks to worker agents. Fails with 409 unless the mission is 'ready'.
Pause Mission — Soft-pause a running mission: no NEW agent runs are dispatched; sessions already running finish their current card. Resume with resume_mission. Use abort_mission for a hard stop.
Resume Mission — Resume a paused mission ('paused' → 'running') and re-dispatch the orchestrator to pick up remaining work (or finish a pending landing).
Abort Mission — HARD-stop a mission: flips it to the terminal 'aborted' state, interrupts and archives every running agent session, and freezes the board. The workspace stays readable as evidence. Not reversible.
Showing the first 25 tools — full list and schemas: Directory API.
MCPBundles probed 105 tools on the live server. The tool list on this page reflects what was discovered at the last refresh — connect your client to see the full set available to your session.
What authentication does Todokan require?
Todokan may require signing in to the provider before tools can run. Connect through MCPBundles or your MCP client and complete any provider login when prompted.
Don't want this listed?
Request removal from the MCPBundles directory. We stop publishing this listing on mcpbundles.com.
MCPBundles is an independent platform built on the open Model Context Protocol standard. MCPBundles is not affiliated with, sponsored by, or endorsed by Anthropic PBC.