Prefer a local server that wraps the full REST API v1? See Workspace MCP. Not sure which surface you need? See Which Taskade MCP do I want?.
The hosted Taskade MCP server connects MCP-compatible AI tools directly to your Taskade workspace.
From your client, browse the workspace and create projects. You can also edit tasks in existing projects.
This is the inbound direction: an AI client reaches into Taskade. For the two outbound paths in the same direction, see Which Taskade MCP do I want?.
Task edits include create, update, complete, move, delete, dates, notes, assignees, and custom fields. These tools arrived in August 2026.
You can manage and prompt AI agents and edit Taskade Genesis app source code.
The toolset mirrors much of Action API v2. It grows with that API.
Availability
The hosted Taskade MCP is generally available on every paid plan. The toolset keeps growing as the Action API v2 surface grows. Read the tool list on this page as a minimum, not as a maximum.
What you can do
- Browse your workspace as a virtual filesystem. It holds projects, agents, automations, uploaded media, and app source code.
- Read structured data. List projects, agents, automations, and tasks. Read a project's full task tree.
- Create and orchestrate. Create projects from Markdown, create, update, and delete agents, attach knowledge, and prompt an agent for its reply.
- Edit your Taskade Genesis app source files (React components, styles, configuration) from your own IDE or AI tool.
What's not included (yet)
The hosted server currently cannot:
- Create, edit, or run automations (flows are listed read-only)
- Upload or delete media
- Register webhooks. Use the Webhook Registration API instead
More tools land as the Action API v2 surface grows.
Requirements
- A Taskade workspace on any paid plan (the Free plan can connect, but workspace access is gated)
- Editor (collaborator) role or above in the workspace to write app source with
write_file. Project and task tools run the same per-action permission checks as the Taskade app. A read-only role can browse, but it cannot modify - An MCP-compatible client such as:
- Claude Desktop
- Claude Code
- Cursor
- Windsurf or Zed
- Any client that supports the Model Context Protocol
Connecting your client
This is a hosted, remote MCP server, so you do not install software.
After you add the URL, an interactive client runs OAuth 2.0 with PKCE for you. The server also advertises dynamic client registration. You do not register an app by hand.
Header-controlled scripts and clients can use a Personal Access Token.
Production endpoint:
https://www.taskade.com/mcp
Taskade publishes OAuth discovery metadata at https://www.taskade.com/.well-known/oauth-protected-resource/mcp for clients that support automatic discovery.
Example: Claude Desktop / Claude Code
Add this to your MCP configuration file:
{
"mcpServers": {
"taskade": {
"type": "http",
"url": "https://www.taskade.com/mcp"
}
}
}
Example: Cursor
In Cursor's MCP settings, add a new server with the URL https://www.taskade.com/mcp.
First-time OAuth authentication
The first time your client connects, Taskade MCP runs a standard OAuth2 flow:
- Let your MCP client discover the Taskade MCP server.
- Follow the browser redirect to Taskade.
- Sign in to Taskade.
- Authorize MCP access to your account.
- Let your client receive and store the token.
Do this once per client. Sign in with the Taskade account that can access the target workspace.
The MCP toolset
The server exposes 32 tools as of September 2026 (the task-level write set landed in August 2026; read the list as a minimum) plus a taskade://getting-started resource, a Markdown quickstart your AI client reads on its own.
The four tables that follow list all 32 by name: 3 + 5 + 14 + 10. Twenty-eight mirror an Action API v2 operation one-for-one. Four are MCP-native and have no v2 twin: inspect_space, read_project, list_automations, and write_file.
Most tools take a spaceId. Call list_spaces first to find it. That list is plan-filtered and paginated, so a page can be empty while more pages remain. Keep paginating until hasNextPage is false.
Browse & app editing
| Tool | What it does |
|---|---|
list_spaces |
List the workspaces you belong to (on plans with MCP access), with your role in each |
inspect_space |
Read files and directories in a workspace's virtual filesystem — app/, projects/, agents/, automations/, media/. Use viewRange: [start, end] to return a specific line range from a file |
write_file |
Create or edit files in your Taskade Genesis app's app/ directory. It can create, replace a substring, insert at a line, or apply a unified diff |
Read workspace data
| Tool | What it does |
|---|---|
list_projects |
Projects in a workspace as structured JSON (id, name) |
read_project |
A project's full contents — task tree with assignees, dates, tags |
list_tasks |
Tasks in a project (id, text, completed, parent) with cursor pagination |
list_agents |
The AI agents in a workspace (id, name) |
list_automations |
The automation flows in a workspace (id, name) |
Task writes
These tools arrived in August 2026. They are the full task-level write set, and they mirror the Action API v2 task operations:
| Tool | What it does |
|---|---|
create_task |
Add a task to a project, optionally under a parent task |
update_task |
Edit a task's content |
complete_task / uncomplete_task |
Check a task off, or reopen it |
move_task |
Reorder a task or move it under a different parent |
set_task_note / delete_task_note |
Set or clear a task's note |
set_task_date / delete_task_date |
Set or clear a task's date |
assign_task / unassign_task |
Assign a project member to a task, or remove them |
set_task_field_value / delete_task_field_value |
Set or clear a custom field value on a task |
delete_task |
Delete a task |
If Google Calendar is connected on an eligible workspace, dates set with set_task_date can sync the same way as dates set in Taskade.
Create & orchestrate
| Tool | What it does |
|---|---|
create_project |
Create a project in a workspace from Markdown |
create_agent |
Create an AI agent (name + role instructions) |
get_agent / update_agent / delete_agent |
Read, update, or delete an agent |
prompt_agent |
Send a one-shot prompt to an agent and get its text reply (runs on the agent's configured tier, consumes credits) |
list_conversations / get_conversation |
Browse an agent's conversation history — get_conversation includes a Markdown transcript |
add_knowledge_project / remove_knowledge_project |
Attach or detach a project as agent knowledge |
If the new agent must read or change project data, pass "enableProjectTool": true to create_agent.
You can change this setting later with update_agent. This setting also enables the required knowledge support.
Tool names track the Action API v2 operations they mirror.
The virtual filesystem
inspect_space presents each workspace as a directory tree:
workspace/
├── app/ # Your Genesis app source (writable via write_file)
├── projects/ # Tasks and documents (use read_project / list_tasks for structured reads)
├── agents/ # AI agent definitions
├── automations/ # Flows (read-only)
└── media/ # Uploaded files (read-only)
You can write files in the app/ directory only. Taskade rejects write_file against projects/, agents/, automations/, or media/. Use the structured tools in the earlier tables to create projects, edit tasks, and manage agents.
Example workflows
Refactor a component with Claude
"Open my 'Landing Page' workspace, look at
app/src/App.tsx, and split the hero section into its own component. Keep the existing styling."
Claude will call list_spaces → inspect_space → write_file (twice) to create the new component file and update App.tsx.
Stand up a project and brief an agent
"Create a project called 'Q3 Launch' with a task per milestone from this outline, attach it to my Research agent as knowledge, then ask the agent to summarize the risks."
The client calls create_project, add_knowledge_project, and prompt_agent. No browser tab needed.
Audit your app for accessibility
"Check every file in
app/src/for missing alt text, ARIA labels, and keyboard handlers. Suggest fixes but do not apply them yet."
The client reads your files with inspect_space. It shows you a report before you approve the edits.
Tips and gotchas
- Always read before you edit. Ask your client to inspect a file first. An edit against a stale view can fail or produce an unexpected result.
- One workspace at a time. Tools take a
spaceId. Point your client at the right workspace before you issue a command. - Publish your changes. MCP updates your Taskade Genesis app source. Publish through Taskade to make the change live on your custom domain.
- Plan gating is per-workspace. A free account can use MCP in a paid workspace where it is a member.
- The same account cannot use MCP in its own Free workspace.
prompt_agentspends credits. It runs the agent for real, exactly like prompting it in the app.
When to use MCP vs. Taskade EVE
Both MCP and Taskade EVE (Taskade's in-product AI assistant) can build and orchestrate. Use whichever fits your workflow:
| Use EVE when… | Use MCP when… |
|---|---|
| You are already inside Taskade | You work in your IDE |
| You want full-scope edits (tasks, flows, app) in one conversation | You script workflows across projects, agents, and app code |
| You build from a prompt end-to-end | You want to use a specific model or tool |
| You prefer a conversational flow | You want precise, file-level control |
If Taskade EVE does not fit your workflow, use MCP with your preferred model, prompts, and keybindings.
Troubleshooting
"MCP access is not available on this workspace's current plan."
Your workspace is on the Free plan. Upgrade to any paid plan, or ask your workspace owner.
"Access denied: you are not a member of this space."
Your account is not a member of the workspace. Make sure that the workspace ID is correct, or request an invitation.
"Access denied: your role … does not have write access. Requires collaborator or above."write_file requires collaborator or above. A read-only role can still browse.
OAuth flow keeps looping.
Sign in to the correct Taskade account in your browser. Then clear the stored client credentials and reconnect.
Feedback
The hosted MCP is in active development, and your feedback shapes what we ship next. If something is missing, broken, or confusing, tell us.
TL;DR: Add https://www.taskade.com/mcp to your MCP client and sign in. Free accounts can authenticate, and MCP works in paid workspaces where they are members. A Free account cannot use MCP in its own Free workspace.
Use any paid plan to create projects, edit tasks, manage and prompt agents, and edit Taskade Genesis app source.
For a local server that wraps REST API v1, see Workspace MCP.