Connect Koji to Cursor: MCP Setup Guide for Product Engineers

Cursor's Model Context Protocol (MCP) support turns the editor into a research-aware coding environment. Connect Koji's MCP server to Cursor and your AI assistant can read live customer interview transcripts, structured answers, themes, quality scores and full study reports, without copy-pasting a single quote. The quickest way in is the Add to Cursor button inside Koji: click it, click Install in Cursor, then approve access in your browser. There is no API key to paste. The MCP connector is included on the Interviews and Enterprise plans.

This guide walks through both setup paths, the most useful Koji and Cursor workflows, and what to check when something doesn't connect.

Why Connect Koji to Cursor

Most teams already use Cursor to write code. Most teams also have customer feedback scattered across Slack threads, Notion docs, and a forgotten interview tool. The MCP integration closes that gap: Cursor can query your Koji workspace the same way it queries the filesystem.

Concretely, this gives product engineers four high-value workflows:

• Context-aware feature work. Before adding a new onboarding step, ask Cursor "what did the last 20 interview participants say about onboarding friction?" and it pulls the actual quotes.
• PRDs grounded in evidence. Generate a one-page spec that cites real interview transcripts instead of speculative bullet points.
• Triage with proof. When grooming the backlog, Cursor can rank tickets against the themes appearing most frequently in published Koji reports.
• Post-launch sense-check. After shipping, ask Cursor to summarise interviews collected since the deploy and flag shifts in sentiment or quality scores.

Survey tools like SurveyMonkey, Typeform, and Qualtrics expect you to log into a separate dashboard to read results. Koji's MCP server brings the data to where engineers already work.

Prerequisites

Before you start, make sure you have:

• A Koji account on the Interviews or Enterprise plan. The MCP connector counts as API access, which is part of those two plans. On Free and Insights, Koji refuses the sign-in. See the Plan Comparison Guide.
• A recent version of Cursor with support for remote MCP servers and OAuth sign-in. If Cursor doesn't offer to sign in when you add the server, update it first.
• A Koji workspace with at least one study and a few completed interviews, so there is something to query.
• A browser for the one-time sign-in.

You don't need an API key. Koji's MCP server accepts only OAuth 2.1 sign-in: Cursor opens your browser, you sign in to Koji and approve the connection, and Cursor stores and refreshes the token for you. API keys from Settings → API Keys are for the REST Headless API, and the MCP server does not accept them.

Option 1: Add Koji from inside Koji (recommended)

1. In Koji, open the Connect your AI row at the foot of the sidebar (or the Use Koji inside ChatGPT or Claude card on your dashboard, if it is still showing).
2. In the Use Koji in your AI app dialog, pick the Cursor tab and click Add to Cursor.
3. Your browser hands off to Cursor, which shows its install prompt with the Koji server filled in. Click Install.
4. Cursor opens a Koji consent page in your browser. Sign in if asked, review the permissions, and approve.

The button opens Cursor's own install link, so it works the same on macOS, Windows and Linux. If Cursor isn't installed on the machine you clicked from, you'll see a web page instead of the install prompt.

Option 2: Add the server to mcp.json by hand

Use this path when you want the config in version control, or when the install link can't reach Cursor (a remote or locked-down machine).

Cursor stores MCP server definitions in a JSON file:

• Global, for every project: ~/.cursor/mcp.json on macOS and Linux, %USERPROFILE%\.cursor\mcp.json on Windows
• Per project: .cursor/mcp.json in the repository root

You can also open the global config from Cursor Settings → MCP. Add Koji to the mcpServers object:

{
  "mcpServers": {
    "koji": {
      "url": "https://www.koji.so/api/mcp/mcp"
    }
  }
}

That is the whole entry. There is no headers block and no secret in the file, so a project-level .cursor/mcp.json is safe to commit. If you already have other MCP servers configured (GitHub, Linear, filesystem), add the koji entry alongside them rather than replacing the file.

Save the file. In Cursor Settings → MCP, the koji server appears with a prompt to sign in or connect. Click it, approve access on the Koji consent page in your browser, and the server's status turns active with its tools listed.

What the consent page asks for

The Koji consent page lists each permission in plain words: reading your studies, interviews and account, and, if the client asks for them, creating and publishing studies, generating and publishing reports, and adding respondents. Write tools only appear in Cursor when you approved the matching permission. Tokens refresh automatically, so you won't be asked to sign in again every session.

Step 3: Verify the Connection

Open a new chat in Cursor (Cmd/Ctrl + L) in Agent mode and type:

List my Koji studies and tell me which one has the most completed interviews.

Cursor should call koji_list_studies and return your study titles, statuses and interview counts. Cursor asks before running each tool unless you've turned on auto-run.

For a deeper smoke test, try:

Summarise the most common themes from my most recent Koji study.

This chains koji_list_studies, koji_get_study and koji_get_interviews. The model picks the right tool for each step, so you don't need to name them. See MCP Tool Reference for the full list.

The Koji Tools You Now Have in Cursor

Once connected, Cursor can call any tool the Koji MCP server exposes:

Read tools: koji_list_studies, koji_get_study, koji_get_interviews, koji_get_transcript, koji_get_account, koji_get_study_data, koji_get_report, koji_export_data

Write tools: koji_create_study, koji_update_brief, koji_publish_study, koji_configure_study, koji_generate_report, koji_publish_report, koji_import_respondents

The read tools are safe to call freely. The write tools change your workspace, so keep Cursor's tool approval on for them. See MCP Authentication and Security.

Recommended Cursor + Koji Workflows

The integration works best when you anchor it to a concrete coding task. Here are five patterns teams use from day one.

1. Writing a feature spec from interview evidence

Open the spec doc and prompt Cursor:

Draft a one-page PRD for a new onboarding checklist. Cite at least
five direct quotes from the most recent onboarding study in Koji,
and group user pain points by theme.

Cursor fetches transcripts and structured answers, then writes the spec with the quotes inline. Structured answers come from Koji's question types (open ended, scale, single choice, multiple choice, ranking and yes/no). See Structured Questions in AI Interviews for how those become chartable data the model can reason over.

2. Ticket grooming with interview evidence

Open your tickets/ folder in Agent mode and prompt:

For each open ticket in this folder, score 1 to 5 how strongly the
last 30 days of Koji interviews support shipping it. Cite the
themes that match.

The model walks each ticket, calls koji_get_study_data for per-question aggregations, and writes the score plus a short rationale into the ticket file.

3. Pre-deploy check

Before merging, ask Cursor to scan recent interviews for complaints about the area you changed. Useful when you ship onboarding, pricing or auth changes, the surfaces participants comment on most.

4. Generating a stakeholder summary

After a study concludes, prompt:

Generate a report for my latest Koji study, then publish it.

This chains koji_generate_report and koji_publish_report. The published report link comes back in the chat, ready to paste into Slack. This is the same flow described in Real-Time Research Insights, driven from your editor.

5. Creating a quick discovery study from a code TODO

Highlight a // TODO: validate this with users comment, open chat, and prompt:

Create a 5-minute Koji study to validate this assumption with
mobile users. Use one open ended question and one scale question
for confidence.

Cursor calls koji_create_study and can publish it with koji_publish_study once you approve, then hands back the interview link.

Troubleshooting

The Add to Cursor button opens a web page. Cursor isn't installed on this machine, or the browser couldn't hand off to it. Use Option 2 instead.

Cursor never opens the sign-in page. Update Cursor, then remove and re-add the server. Make sure your browser isn't blocking pop-ups from koji.so.

Sign-in completes but the server stays disconnected. Your account isn't on a plan with API access. The MCP connector needs the Interviews or Enterprise plan.

Authentication errors after setup worked before. Check your mcp.json for an old headers block with Authorization: Bearer .... Older guides suggested an API key there; the MCP server doesn't accept API keys. Delete the headers block and sign in again.

Write tools are missing. You approved read access only. Disconnect the server in Cursor Settings → MCP and connect again, approving the write permissions.

"Tool not found" errors. Toggle the server off and on in Cursor Settings → MCP, and confirm the JSON is valid.

For more, see the MCP Troubleshooting Guide.

How This Compares to Pasting Transcripts

Some teams still copy interview snippets into Cursor by hand. That works for a single decision but breaks down at scale: stale data, missing quality scores, and no way to filter by theme or sentiment.

The MCP integration fixes all three. Cursor sees the same data your researchers see in Koji, limited to the studies your account can access, with each interview's quality score attached so the model can weigh strong interviews over thin ones. See Understanding Quality Scores.

Related Resources

• Koji MCP Integration Overview
• Connect Koji to Claude (Setup Guide)
• Connect Koji to VS Code (Copilot Chat)
• MCP Tool Reference
• MCP Authentication and Security
• MCP Best Practices
• Structured Questions in AI Interviews

Further reading on the blog

• B2B Customer Research: The Complete Guide for Product Teams (2026) — B2B customer research is harder than B2C — you are navigating buying groups of 10+ stakeholders, gatekeepers, and enterprise procurement cyc
• Best Product Discovery Tools in 2026: The Complete Buyer's Guide — Product discovery is no longer a one-off pre-launch phase — it is a continuous loop. The best teams in 2026 are running weekly customer inte
• Concept Testing: The Complete Guide for Product Teams (2026) — Concept testing validates whether your idea is worth building before you build it. This guide covers methods, question templates, analysis a

<!-- further-reading:blog -->

---

<!-- crosslink:dev-hub -->

Related: other ways to connect

The Developer Hub is the single starting point for every way to connect to Koji: the MCP server (this section), the REST Headless API, and the embed widget. For a code integration, see API Authentication and Starting Interviews via API.