Product management MCP servers

A server here product roadmaps, feature requests or product decisions.

Not here: project scheduling.

565servers
409 remote156 localTagged by Odel's AI, which can make mistakes
By GitHub stars, at most two per publisher
  1. Number 1: Idea Reality@mnemox-ai/idea-reality-mcpPre-build reality check. Scans GitHub, HN, npm, PyPI, Product Hunt — returns 0-100 signal.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.822 GitHub stars.
  2. Number 2: AsDecided@asdecided/coreRead-only access to deterministic engineering decisions stored in a local repository.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from Docker.296 GitHub stars.
  3. Number 3: Reddit research@king-of-the-grackles/reddit-research-mcpTurn Reddit's chaos into structured insights with full citations - MCP server for Reddit research5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.254 GitHub stars.

Best fit

220 made for this, ranked by starsRemoteLocal
  1. Alive Analysis @with-geun/alive-analysisQuery alive-analysis history from any AI client. Tools: list, get, search, dashboard export.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.41 GitHub stars.
  2. Linksee Memory @michielinksee/linksee-memoryHand a project over with the reasons attached. Local-first cross-agent memory MCP + drift detection.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.14 GitHub stars.
  3. Nauro @nauro/nauroDecision system for software built with AI3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.11 GitHub stars.
  4. Lockstep @lockstep-team-agent/lockstepDecision memory for AI coding agents: captures decisions once, briefs every agent before it acts.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.10 GitHub stars.
  5. Weckr @useweckr/weckrAsk which of your customers cost more in LLM calls than they pay, per user and per feature.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.8 GitHub stars.
  6. Aha.io @cedricziel/aha-mcpSearch, read and update Aha.io ideas, features, epics, releases, goals and comments5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.6 GitHub stars.
  7. Product CLI @hafeok/product-cliKnowledge graph CLI for managing features, ADRs, and test criteria as MCP-discoverable artifacts.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from MCPB.6 GitHub stars.
  8. PRD Spec Generator @cdeust/prd-spec-generatorTurns a feature description into a verified 9-file PRD — multi-judge consensus, 17 MCP tools.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from MCPB.5 GitHub stars.
  9. Human Standards @aklodhi98/human-standardsRead-only Human Standards guidance for accessible, usable interface design.2% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.4 GitHub stars.
  10. AppMetrica @pavellunev/appmetrica-mcpAppMetrica (Yandex) analytics: reports, funnels, segments, dashboards, logs export, pushNot assessed yet.Runs on your machine, installed from npm.3 GitHub stars.
  11. App Store Operator @meyusufdemirci/app-store-operatorFree App Store competitive intelligence for Claude — rival downloads, revenue, and ASO keywords.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.3 GitHub stars.
  12. Pathmodeio MCP @pathmodeio/mcp-serverDeterministic intent preflight before your agent builds: six gates, keyless, no model call.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.3 GitHub stars.
  13. ProductPlan MCP Server @olgasafonova/productplan-mcp-serverTalk to roadmaps with AI. Query and manage bars, OKRs, ideas, and launches.4% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from Docker.3 GitHub stars.
  14. rewelo @sebs/reweloPrioritise a backlog with the relative weight method: score, rank, simulate and plan tickets4% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.2 GitHub stars.
  15. Cli @aligndottech/cliCross-tool decision memory for AI coding agents: search your decision graph, check changes over MCP.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.2 GitHub stars.
  16. Story Mapper @tomatap/storymapperUser story map + kanban with a live browser UI; AI agents and humans plan the same backlog via MCP.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.2 GitHub stars.
  17. Christensen @m2ai-mcp-servers/mcp-christensenClayton Christensen persona agent for strategic advisory using disruption theory5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.2 GitHub stars.
  18. Kontra @bluwork/kontra-mcpThe Tenth Man Protocol — structured contrarian analysis for decision-makers via MCP4% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.2 GitHub stars.
  19. Grill @mtangoz/grillA second opinion from an unrelated AI model before you commit to a decision.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from MCPB.1 GitHub stars.
  20. DemandScope @baobabcat/demandscopeLive demand signals for idea validation: npm/PyPI downloads, GitHub + HN attention trends.4% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from MCPB.1 GitHub stars.
  21. Dotplot @brownglasses/dotplot-mcpProduct analytics for user flows: aha moments, retention, funnels, per-user dot plots. YC method.4% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.1 GitHub stars.

Also touches on Product management

95 where it is one part of a bigger job