Productboard MCP

What it does

The Productboard MCP server lets AI tools such as Claude, ChatGPT, Claude Code, Codex, Cursor, Lovable, and Claude Design read and update data in a Productboard workspace. Users give the agent plain-language prompts, and it can search and edit features, initiatives, releases, objectives, key results, companies, users, feedback, and documents.

People use it to pull current product specifications into coding agents, update prioritization fields such as value, risk, effort, and confidence, move features between initiatives and releases, and leave comments on specs. Access follows each member's workspace role and permissions, so an agent only reaches what the signed-in user can already see. Every write is captured in version history.

The server suits product managers, product operations teams, and engineers working with coding agents. Workspace admins enable MCP in Settings and choose which roles may connect, and each member signs in once with a Productboard login. Setup steps differ by agent, and agents may show IDs rather than names unless asked otherwise.

What it can do

• Search and read features, initiatives, documents, and specifications from a connected agent
• Create and update product entities such as features, initiatives, releases, objectives, and key results
• Update prioritization fields such as value, risk, impact, confidence, and effort
• Add features to initiatives, move initiatives between releases, and create key results under objectives
• Pull product specifications into coding agents and leave comments back on the original spec
• Track every agent write in version history

Works with

• Claude
• ChatGPT
• Claude Code
• Codex
• Cursor
• Lovable
• Claude Design

Before you choose

• Each member connects with their own login and sees only what their role allows; admins can revoke all connections with Disconnect all members.
• Written content from agents is real workspace data and notifies teammates, and changes are captured in version history.
• Setup steps differ across Claude, ChatGPT, Claude Code, Codex, and Cursor, so follow the instructions for the agent in use.

Listed under

• Coding
• MCP and agent tools

Get started

A workspace admin enables MCP under Settings → Integrations → MCP server; members then connect their own agents with browser sign-in.

1. Admin enables MCP for the workspace under Settings → Integrations → MCP server and selects which roles can connect.
2. Member adds the connector in their agent using the server URL https://mcp.productboard.com, choosing HTTP and OAuth where prompted.
3. Member signs in with their Productboard login once; the connection then respects their workspace role and permissions.

Business requires requesting a demo and Enterprise requires contacting sales. Spark usage draws on monthly AI credits, with Business including 500 credits per maker each month.

Productboard MCP doesn’t publish prices. You get a quote by contacting sales. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Productboard MCP Server ↗www.productboard.com
• Productboard Plans & Pricing ↗www.productboard.com
• The Agentic Product Management System | Productboard ↗www.productboard.com
• The AI Agent Built for Product Managers | Productboard Spark ↗www.productboard.com
• API 2.0 | Productboard ↗www.productboard.com
• Connect a coding agent to Spark via MCP – Productboard Support ↗support.productboard.com

Company website

productboard.com: product, pricing and policy pages

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.