MCP

Overview

DolphinQ connects to your AI assistant via MCP — so you can create stories, wire dependencies, and visualize timelines entirely from chat.

What MCP Enables

With MCP connected, your AI assistant can:

  • Create projects and components, and create or update stories one at a time or in batches of up to 50
  • Fetch many stories at once as rich cards (status, risk, dependencies, counts)
  • Wire dependencies between stories automatically
  • Query project status, blockers, and critical path
  • Analyze stories and save the assessment (with history) as a shareable page
  • Attach artifacts — Drive, Notion, Figma, or any URL — to stories
  • List members, components, and assigned work
  • Add tagged, markdown-formatted comments and progress updates to stories
  • Populate Gantt timelines and dependency DAGs in real time

How It Works

DolphinQ runs a Streamable HTTP MCP server. When your AI connects:

  1. It discovers the full set of tools for workspace operations
  2. You describe what you need in natural language
  3. Your AI calls DolphinQ tools to create stories, set dependencies, check status
  4. Everything appears instantly in your workspace — Kanban, Gantt, and DAG

Authentication

DolphinQ supports two authentication methods:

  • OAuth (recommended) — Just add the MCP server URL. Your AI opens your browser for a one-time sign-in. You stay authenticated for up to 90 days with automatic token refresh.
  • Token-based — Generate an encrypted token (dlpq_ prefix) in Workspace Settings → Integrations and send it as an Authorization: Bearer header.

Both methods scope every MCP action to your identity and workspace.

Next Steps

Tool Reference →Every MCP tool with parameters and examples