TL;DR
Cookiy is an agentic user-research platform that operates as a Model Context Protocol (MCP) server for AI coding agents like Claude Code and Cursor. It allows developers to delegate end-to-end qualitative research—from study design and participant recruitment to conducting AI-moderated voice interviews and generating reports—directly from their IDE.
What Users Actually Pay
No user-reported pricing yet.
Our Take
Cookiy represents the shift from 'AI as an assistant' to 'AI as an autonomous operator.' While most MCP servers focus on fetching documentation or executing code, Cookiy provides a high-level workflow for product discovery and user validation. It is uniquely positioned for 'founder-developers' who need to validate features quickly without leaving their development environment. The product's strength lies in its integration of a massive participant panel with sophisticated AI-led interviewing, significantly reducing the cost and time of qualitative research. However, it is heavily reliant on the relatively new MCP ecosystem, which can be prone to 'context-window bloat' and setup friction for non-technical stakeholders. It is best suited for early-stage startups and product teams using AI-first IDEs (Cursor, Windsurf) who want to replace expensive research agencies with scalable, agentic workflows. For enterprise teams, the SOC2 and GDPR compliance layers make it a viable alternative to legacy tools like UserTesting.
Similar Products
Pros
- + Seamless IDE Integration: Operates natively within Claude Code, Cursor, and Codex via a simple 'npx' installation.
- + End-to-End Automation: Handles everything from the initial 'plain-English' research goal to the final insight report.
- + Scalable Qualitative Research: Can conduct hundreds of simultaneous AI-moderated interviews, a task that would take human researchers weeks.
- + Generous Free Tier: The Starter plan includes 50 AI interviews per month, lowering the barrier for entry for solo developers.
- + High-Fidelity AI Moderation: Uses agentic voice AI that can 'read the room' and pivot questions based on user sentiment.
Cons
- - Ecosystem Dependency: Its performance is tied to the stability of the Model Context Protocol (MCP) and the specific AI client being used.
- - Context Consumption: Like many complex MCP servers, it can consume a significant portion of an LLM's context window during long planning sessions.
- - Learning Curve: Users must learn specific tool-calling syntax (e.g., cookiy_introduce) to orchestrate complex multi-step studies effectively.
- - Niche Focus: Primarily valuable for user research; developers looking for general coding skills may find it too specialized.
Sentiment Analysis
Sentiment has improved since last capture. Sentiment has significantly improved (from 0.30 to 0.78) following the official launch of its MCP server and its $7M funding announcement. Developers are particularly impressed with the product's 'agentic' nature, though some platform-wide skepticism regarding the Model Context Protocol's efficiency remains a minor theme.
Sentiment Over Time
By Source
12 mentions
Sample quotes (2)
- "One standout MCP is Cookiy... it actually turns the agent into a researcher instead of just a glorified search tool."
- "I'm skeptical about MCP context eating, but the automation of participant recruitment is a game changer for solo devs."
45 mentions
Sample quotes (2)
- "The installer auto-detects clients perfectly. Easiest MCP setup I've seen so far."
- "Finally a practical use case for agentic workflows beyond just writing code."
15 mentions
Sample quotes (2)
- "Cookiy AI just raised $7M to build the first agentic research operator. This is what the future of product discovery looks like."
- "Check out the Cookiy MCP—it's like having a UX researcher living in your VS Code."
5 mentions
Sample quotes (1)
- "Officially listed in the MCP registry. Essential for anyone building user-centric products with AI agents."
Agent Readiness
59/100Cookiy is a 'Tier 1' agent-ready product. It is built specifically for AI agents using the Model Context Protocol (MCP), meaning it requires no 'wrapper' to be used by autonomous systems. It features a robust CLI (npx cookiy-mcp), standard OAuth 2.1 authentication, and a well-defined tool contract that guides agents through state-machine-aware research workflows.
Last checked Mar 29, 2026
MCP Integrations
1 serverAI user research via study creation, interviews, recruitment, and reports.
Last checked Mar 18, 2026
Compare With
Reviews
No reviews yet. Be the first to review ai.cookiy/cookiy!