Cadylo MCP

Your workspace. One URL away.

Connect Codex, Claude or another AI agent to Cadylo. Plan projects, update issues and manage your work from the conversation.
Add the remote MCP URL, sign in to Cadylo and choose your workspace. No connector download or local server to maintain.

Work across Cadylo.

01

Issues and collaboration

Create, update, assign and delete issues. Manage comments, labels, parent issues and dependencies.

Create the launch tasks and link their dependencies.

02

Projects and planning

Manage projects, milestones, cycles and progress updates, including automatic cycles.

Plan the next cycle and add a milestone for the release.

03

Workspace and teams

Discover your workspaces, members and roles. Create and update teams and product groups.

Show my teams and the projects that need attention.

04

Documents and fields

Read and edit versioned documents. Manage typed custom fields on issues and projects.

Update the release brief with the decisions in this conversation.

05

Workflows and automation

Manage workflow states, saved views, issue templates, recurring schedules and triage rules.

Create a recurring issue template for our weekly review.

06

Progress and notifications

Read analytics, delivery evidence and integration status. Manage your notifications and subscriptions.

Summarize what shipped this week and the work still blocked.

The MCP exposes the operations listed here. Account deletion, billing, provider credentials, file uploads and workspace ownership transfer remain in the app. Your agent discovers the current tool list when it connects.

Connection guide

Add. Authorize. Work.

Use the URL directly, a CLI command or the JSON configuration. Authentication opens Cadylo so you can approve the assistant with your own account.

  1. 1. Add the Cadylo MCP URL

    In your assistant, choose a remote MCP server or connector and paste this URL. Cadylo runs the server for you.

    https://cadylo.app/mcp
    Remote MCP connection
    codex mcp add cadylo --url https://cadylo.app/mcp
  2. 2. Sign in and authorize your assistant

    Follow the sign-in prompt, use your Cadylo account and review the requested access. Your assistant receives its own connection; you do not need to copy a personal API token.

    In Codex, run codex mcp login cadylo if authentication has not opened. In Claude Code, open /mcp and authenticate Cadylo. Other clients must support remote MCP over HTTP and OAuth.

  3. 3. Choose your workspace and start working

    Ask your assistant to list your Cadylo workspaces, then tell it which one to use. Your current roles determine what it can see and change.

    “Show my Cadylo workspaces, then list the issues in progress.”

    Revoke an assistant in Settings → Developer & MCP → Authorized assistants.

Your account. Your permissions.

Your assistant can work across the workspaces you can access. Private resources stay subject to your roles, and each request checks the current access rules. Review deletions and broad changes with your assistant before it performs them.

Stay in control of the connection

Revoke an assistant in Developer & MCP. OAuth manages the connection credentials, so the setup examples contain no secrets. The content you request is sent to your chosen MCP client and AI provider.

Connection help

My agent is asking me to authenticate

Open its Cadylo connection and follow the OAuth sign-in prompt. Codex also supports codex mcp login cadylo; Claude Code provides authentication through /mcp.

An operation is unavailable or denied

Check the selected workspace and your permissions in Cadylo. Reconnect to refresh the tool list. A JSON-only client must also support remote HTTP transport and OAuth authentication.