MCPs for Product Managers

Official MCP server providing standardized interface for Linear data access. Find, create, and update issues, projects, and comments. Supports OAuth 2.1 with dynamic client registration.

PM use case: Saves ~1.5 hrs/week. Ask 'what's blocking the checkout redesign?' and get structured answers in 30 seconds — no clicking through Linear filters. Best use case is standup prep: ask for all high-priority bugs assigned to a specific team this sprint and get a structured summary instantly. Also great for daily triage and creating tickets directly from PRD specs.

Official GitHub MCP server connecting AI tools to GitHub's platform. Read repositories and code files, manage issues and PRs, analyze code, monitor GitHub Actions, and automate workflows through natural language.

PM use case: Pull implementation details into PRDs before writing specs ('show me how the payment flow is implemented'). Read a pull request diff and generate user-facing release notes automatically. Create GitHub issues directly from a planning doc. PMs who work closely with engineering get the most value — understanding technical constraints before committing to timelines.

Official hosted server giving AI tools secure access to Notion workspaces. Read and write pages, create documentation, search content, manage tasks, build reports, and plan campaigns directly within Notion.

PM use case: Saves 3–4 hrs/week. Real example: Claude extracted 6 user stories across 2 personas, flagged 3 dependencies that had been missed, and created properly tagged sprint database entries — all from a 5-page PRD in 11 minutes. Also great for searching product specs while writing code, and syncing meeting notes with implementation details. The hosted version uses OAuth with one-click setup.

Official cloud-based bridge to Atlassian Cloud (Jira, Confluence, Compass). Summarize and search content, create and update issues or pages via natural language, automate ticket generation from meeting notes.

PM use case: Best for teams deep in the Atlassian ecosystem: bridges both Jira and Confluence in one server. Query Jira for 'Ready for Grooming' issues, summarize each ticket with context, then post a structured summary to a Slack channel. Also great for backlog grooming: ask for all epics tagged for this quarter and get a prioritized view. Requires CONFLUENCE_URL, JIRA_URL, and separate API tokens for each.

Official MCP server for Asana Work Graph access. Create and manage tasks/projects through natural language, generate reports and summaries, analyze project data with AI-powered insights.

PM use case: Find incomplete tasks with natural language ('show my overdue tasks in the mobile redesign project'), create tasks in projects without opening Asana, generate weekly status reports, and analyze project timelines and dependencies. Works directly in Claude and ChatGPT without additional setup beyond authentication.

Community-maintained MCP server for Slack workspaces (originally official, now by Zencoder). List channels, post messages, reply to threads, add reactions, get channel history, and manage users.

PM use case: Saves ~45 min/week. The killer use case isn't posting messages — it's mining conversation history. Ask Claude to find what the engineering team decided about your authentication approach back in October, or pull out key decisions from a long thread about your pricing model. Also useful for drafting launch updates based on latest ticket statuses and posting to relevant channels.

Integrates PostHog product analytics with Claude Code. Query insights, list projects, create annotations for important events, check feature flag status, and search PostHog documentation.

PM use case: Saves ~2 hrs/week on meeting prep. Before a product review, ask 'What's our 7-day retention for users who completed onboarding last month?' and get the answer in 15 seconds — no dashboard prep, no building slides. During a stakeholder meeting, answer unexpected questions about conversion rates for specific user segments on the spot. Requires PERSONAL_API_KEY and POSTHOG_REGION ("us" or "eu").

Web search with citations, deep research, and step-by-step reasoning — powered by Perplexity AI. Returns sourced answers you can verify and share with stakeholders.

PM use case: Run competitive research without leaving Claude Code. Ask it to find the latest pricing changes for three competitors, summarize each with source URLs, and save the analysis as a markdown file. The citations let you verify every claim before sharing with stakeholders. Best for market research, competitive intelligence, and any work where you need sourced answers rather than AI-generated summaries.

Scrape, crawl, and extract structured data from any website. Handles JavaScript rendering and modern SPAs — unlike basic scraping, it can read dynamically loaded content.

PM use case: Monitor competitor landing pages for messaging and pricing changes. Firecrawl renders JavaScript-heavy pages so you can extract content from modern SPAs. Set up a weekly workflow that crawls competitor pages and diffs against last week's snapshot — useful for catching pricing updates, feature announcements, and positioning shifts before your team does.

Semantic search across the web — finds content by meaning, not just keywords. Surfaces high-quality, expert-level sources that keyword search typically misses.

PM use case: Search for 'B2B SaaS companies that recently changed their pricing model' and get results ranked by semantic relevance, not SEO keywords. Useful for market research, finding case studies, and discovering how other PMs solved similar problems. Complements Perplexity MCP — use Perplexity for current events and citations, Exa for deep semantic discovery of technical and domain-specific content.

Query databases, manage tables, and run SQL directly from Claude Code. Works with any Supabase project and handles authentication, schema inspection, and data analysis.

PM use case: Answer ad-hoc product questions from your actual database: 'How many users signed up last month?', 'What's the most common plan type?', 'Which features do churned users never activate?' Claude writes and runs the SQL for you — useful for PMs who want data answers without waiting for an analyst or learning SQL themselves.

Manage Vercel deployments, environment variables, domains, and function logs from Claude Code. Read-only access in initial release, with project and deployment inspection.

PM use case: Check deployment status after a release without asking engineering. Read production logs to understand error rates before a stakeholder meeting. Most useful for PMs who ship internal tools or manage their own sites — keeps you informed about infrastructure state without interrupting developers for status updates.

Keyword research, SERP analysis, backlink data, and SEO metrics on demand — the same data as Ahrefs or SEMrush, queried directly from Claude Code.

PM use case: Research keyword volumes before writing a feature page or blog post. Analyze competitor SERP rankings for your target terms. For PMs responsible for growth or content, DataForSEO MCP provides programmatic access to search data — useful for validating content ideas, tracking ranking changes, and understanding what queries your users are actually searching before you build.

Search, read, and create Google Docs, Sheets, and Slides from Claude Code. Works with the full Google Workspace suite including shared drives.

PM use case: Import a spreadsheet of user survey responses and ask Claude Code to summarize key themes and sentiment. Read a Google Doc containing stakeholder feedback, then generate structured action items. Useful for PMs whose organization lives in Google Workspace rather than Notion — same workflow benefits, different document backend.

Connects Claude Code to the LaunchDarkly feature flag platform. List projects and environments, check flag status, search feature flags, and monitor rollout state across environments.

PM use case: Ask 'Is the new onboarding flow live in production?' or 'Which flags are currently enabled for the checkout experiment?' without opening the LaunchDarkly dashboard. Best for teams running 10+ simultaneous experiments where checking flag status across staging and production is a regular interruption. Requires LD_API_KEY with read permissions.