Connecting AI to ManicTime Cloud

ManicTime Cloud supports MCP (Model Context Protocol), so AI tools can connect directly to your workspace and query activity data without going through the local desktop client.

If you want to connect an AI tool to the desktop app running on your computer, see Connecting AI to the desktop app.

If you want to connect an AI tool to your own ManicTime Server, see Connecting AI to ManicTime Server.

Why use ManicTime Cloud MCP

Use ManicTime Cloud MCP when:

  • your data is already in ManicTime Cloud
  • your AI tool supports remote MCP servers with OAuth
  • you want the AI tool to connect directly to your workspace
  • you need to query data for users or teams you are allowed to see

Your ManicTime Cloud MCP server URL looks like this:

https://cloud.manictime.com/<your-cloud-id>/mcp

The <your-cloud-id> part is unique to your workspace. You can find it in the address bar when you are signed in to ManicTime Cloud.

Permissions and user data

The data available through MCP follows the permissions of the signed-in user.

  • Regular users can query their own data.
  • Administrators can query data for other users or everyone.
  • Users with Content reader permissions can query data they are allowed to read, such as all users or users in their team.
  • Users with Content writer permissions can query and write data they are allowed to manage.

When asking about other people or teams, tell the AI which user, team, or scope to use. The MCP tools will only return data that account is allowed to access.

What you need

Before you start, you need:

  • a ManicTime Cloud account
  • access to the Developer page
  • an AI tool that supports remote MCP servers with OAuth

In ManicTime Cloud, you can open the Developer page here:

  • Single-user accounts: bottom-left menu → My profile → Developer
  • Team accounts: bottom-left menu → Users and subscriptions → Developer

Connect ChatGPT

In ChatGPT, go to:

Settings → Apps → Create app

First you need to enable Developer mode.

Then click on Create app. Fill in:

  • Name: ManicTime
  • MCP Server URL: https://cloud.manictime.com/<your-cloud-id>/mcp
  • Authentication: OAuth

Under advanced authentication settings, ChatGPT will show a callback URL.

Then in ManicTime Cloud:

  1. Open the Developer page.
  2. Create a new client.
  3. Paste the callback URL provided by ChatGPT.
  4. Save the client.
  5. Copy the generated Client ID and Client Secret.

Go back to ChatGPT, paste the Client ID and Client Secret, then click Create.

After that, ChatGPT should complete OAuth and connect to your ManicTime MCP server.

Connect Claude

Claude also supports remote MCP servers with OAuth.

To add a custom connector, go to:

https://claude.ai/customize/connectors

Then:

  1. Click +, then Add Custom connector.

  2. Enter:

    • Name: ManicTime
    • MCP Server URL: https://cloud.manictime.com/<your-cloud-id>/mcp

  3. In ManicTime Cloud, open the Developer page and create a new client.

  4. For callback URL, use:

    https://claude.ai/api/mcp/auth_callback

  5. Save the client.

  6. Copy the generated Client ID and Client Secret.

  7. Go back to Claude and paste the Client ID and Client Secret.

  8. Finish creating the connector.

Connect other AI tools

The same pattern works for any AI tool that supports:

  • a remote MCP server URL
  • OAuth authentication
  • a custom callback URL

The setup is always the same:

  1. Start creating a new MCP connection in the AI tool.
  2. Enter your ManicTime Cloud MCP URL.
  3. Choose OAuth.
  4. Copy the callback URL the tool provides.
  5. In ManicTime Cloud, create a new client with that callback URL.
  6. Paste the client ID and client secret back into the AI tool.
  7. Complete the OAuth flow.

Check that it works

Once the connection is active, try a simple prompt first:

  • "Are ManicTime tools available to you?"

Then you can try questions like:

  • "How much time did I work yesterday?"
  • "Which projects did I spend the most time on this week?"
  • "Show me all tagged and untagged work from Monday."
  • "Summarize my support work for today."
  • "Show total time for everyone on my team yesterday."
  • "Which projects did Sara work on this week?"

Depending on the MCP client and the permissions you allow, the AI may also be able to write data back, for example by saving tags.

For the full list, see Available AI (MCP) tools.