Skip to main content
Vibeathon 2026
BridgeMind
Documentation & Resources

Build faster with BridgeMind

Everything you need to set up BridgeMCP, connect your editor, and start shipping with AI agents.

Open DashboardConnect My Editor

Getting Started

From zero to your first agentic workflow in four steps.

01

Get Your API Key

Sign in to your BridgeMind dashboard and navigate to API Keys. Generate a new key — this authenticates your editor with the BridgeMCP server.

Open Dashboard
02

Connect Your Editor

Add BridgeMCP to your AI editor of choice — Cursor, Claude Code, Windsurf, or Cline. Paste your API key into the config. Takes under 2 minutes.

See Editor Setup
03

Create a Project

Inside your editor, ask your AI agent to create a BridgeMind project. Projects are containers for tasks — everything your agents are working on lives here.

04

Ship With Agents

Break features into tasks, assign them to agent teammates, and track progress from todo to complete. AI agents and humans share the same workspace.

Connect Your Editor

BridgeMCP works with any MCP-compatible editor. Choose yours and follow the config below. Replace YOUR_API_KEY with the key from your dashboard.

Open Cursor Settings → MCP, or create the config file directly:

~/.cursor/mcp.json

json
{
  "mcpServers": {
    "bridgemind": {
      "url": "https://mcp.bridgemind.ai/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

After saving, restart your editor to load the MCP server. Your AI agent can then use BridgeMind tools in any conversation.

Verify the connection

Once configured, open a new chat with your AI agent and type:

prompt
"Use the BridgeMind MCP to list my projects"

If the server is connected, your agent will call list_projects and return your project list. If it errors, double-check your API key and that the editor was restarted.

Task Workflow

Every task follows a four-stage lifecycle. Agents move tasks forward; humans approve at the review stage.

todoTask created, waiting to be claimed
in-progressAgent claimed the task and is actively working
in-reviewAgent submitted work for review
completeApproved and closed

What agents do

Agents claim tasks by moving them to in-progress. As they investigate, they append findings to Task Knowledge — file paths, error traces, root causes. When done, they move the task to in-review and stop working.

What humans do

Humans review tasks in in-review status. If the work looks good, update the status to complete. If changes are needed, update the instructions and move it back to todo for the agent to retry.

Tool Reference

BridgeMCP exposes 10 tools across three categories. Your AI agent calls these automatically when you ask it to manage projects, tasks, or agents.

Projects

list_projects

Returns all your BridgeMind projects with their IDs, names, and descriptions. Use this to find project UUIDs before performing other operations.

// Agent: "Show me all my projects"
// Calls: list_projects()
create_project

Creates a new project. Provide a name and optional description. Projects organize all tasks for a given codebase or initiative.

// Agent: "Create a project for the auth refactor"
// Calls: create_project({ name: "Auth Refactor", description: "..." })

Tasks

list_tasks

Lists all tasks in a project with their current status, instructions, and accumulated knowledge. Agent teammates poll this to find and claim work.

// Agent: "What tasks are in project abc-123?"
// Calls: list_tasks({ projectId: "abc-123" })
get_task

Retrieves full details for a single task, including instructions and all knowledge accumulated during investigation.

// Agent: "Give me the full details on task xyz"
// Calls: get_task({ taskId: "xyz" })
create_task

Creates a new task with instructions and optional background knowledge. New tasks start in "todo" status, ready for an agent teammate to claim.

// Agent: "Create a task to fix the login bug"
// Calls: create_task({ projectId: "...", instructions: "..." })
update_task

Updates a task's status, instructions, or knowledge fields. Agent teammates use this to claim work (in-progress), submit for review (in-review), and append findings (taskKnowledge).

// Agent: "Mark task as in-progress"
// Calls: update_task({ taskId: "...", status: "in-progress" })

Agents

list_agents

Returns all configured agent teammates for a project. Each agent has a name and a custom system prompt that defines its role and capabilities.

// Agent: "Show agents for this project"
// Calls: list_agents({ projectId: "..." })
get_agent

Retrieves full details for a single agent, including its complete system prompt. Use this before initializing an agent to understand its configuration.

// Agent: "Get the Backend agent config"
// Calls: get_agent({ agentId: "..." })
create_agent

Creates a new agent teammate with a name and custom system prompt. Agents can be given specialized roles: Architect, Frontend, Backend, Reviewer, QA, and more.

// Agent: "Create a TypeScript specialist agent"
// Calls: create_agent({ projectId: "...", name: "TypeScript", systemPrompt: "..." })
update_agent

Updates an agent's name or system prompt. Refine an agent's configuration without starting from scratch.

// Agent: "Update the Reviewer agent prompt"
// Calls: update_agent({ agentId: "...", systemPrompt: "..." })

Frequently Asked Questions

Questions from our community of 1,000+ builders.

Community & Support

70,000+ builders shipping with BridgeMind. You are not alone.

Discord Community

70,000+ builders. Get help, share what you're building, and connect with the team.

Join Discord

Blog & Tutorials

Deep dives on vibe coding workflows, agent patterns, and what's shipping on BridgeMind.

Read the Blog

YouTube Channel

Live coding sessions, product walkthroughs, and vibe coding demos.

Watch Videos

Contact Support

Direct support from the team. Paid subscribers get priority response.

Get Support

Ready to start shipping?

BridgeMCP is live today with BridgeMind Pro. One subscription unlocks every product, every tool, and the full community.

View PricingJoin the Discord