What is the Askable MCP server?
The Askable MCP server lets you query your research data from inside the AI tools you already use, including Claude, ChatGPT, Microsoft Copilot, Cursor, Figma Make, Lovable, and others.
Ask a question in natural language and get back answers grounded in real participant quotes, with links to the original highlights and recordings in Askable. No more switching tabs to check your research. No more confident-sounding AI summaries you can't verify.
What you can do
Search your research. Ask questions like "What are users saying about onboarding friction?" and get back direct participant quotes with source links.
Browse studies. See what research is available in your workspace and explore study details.
Read transcripts. Pull up full interview transcripts with speaker labels, timestamps, and participant context.
Explore participants. View demographic details, screener responses, and session recordings for participants linked to your search results.
Filter results. Narrow your search by age, gender, location, time range, or specific studies directly in your questions.
Query across teams. If you have access to multiple teams or industry streams, you can search across them. The server switches context as needed and brings back a combined response.
What you'll need
An Askable account on a team with AI-enabled tracks. If you have access to Ask AI, you have access to this.
An AI tool that supports the Model Context Protocol (MCP).
Setup instructions for MCP-compatible tools
Any tool that supports the Model Context Protocol can connect to Askable. Setup happens in two stages.
First, an admin of the MCP-compatible tool adds the connector. They'll find this in the tool's MCP or connector settings, where they enter:
Connector Icon: Download here
Name:
AskableDescription:
Your research, where you workAuthentication: OAuth
Once the connector is in place, anyone with an Askable account can sign in to use it.
Claude
Setup happens in two stages. An Owner adds the connector once for the whole organisation. Each member then authenticates individually when they connect.
Part 1: Organisation Configuration
Sign in as an Owner or Primary Owner.
Go to Organisation settings → Connectors (https://claude.ai/admin-settings/connectors)
Click Add
Hover over Custom, then select Web
Paste the MCP Server URL:
https://mcp.askable.com/mcpClick Add
The connector is now available to everyone in your organisation.
We are in the process of becoming an approved connector with Anthropic. This documentation will be updated once that's completed.
Part 2: Askable User Setup
In Claude, go to Customize → Connectors (https://claude.ai/customize/connectors)
Find Askable in the list. It will be labelled Custom.
Click Connect
Authenticate in Askable
Open a new chat
Click the + at the bottom left of the chat, select Connectors, and toggle Askable on
Start chatting. Claude will invoke Askable's tools when relevant, or you can ask it to use Askable directly.
ChatGPT
Before starting, invite the ChatGPT Workspace Admin to the Askable team and ask them to join. They will need to Authenticate in Askable when setting up the App connector.
Part 1: Workspace Configuration
Ensure you're logged in as a Workspace Admin and using the Workspace profile.
In ChatGPT, go to Profile → Workspace settings → Apps (https://chatgpt.com/admin/ca) → + Create
If you haven't already, click Enable Developer Mode.
Paste the MCP Server URL:
https://mcp.askable.com/mcpClick Create
Authenticate in Askable
Click Publish
Review the potential risks and click Publish
We are in the process of becoming an approved app in ChatGPT. This documentation will be updated once that's completed.
Part 2: Askable User Setup
In ChatGPT, go to Apps in the left menu
Click on Askable
Click Connect
Continue to Askable
Authenticate in Askable
Click New Chat
Type "@" and select Askable to start using
Figma Make
Setup happens in two stages. An organisation admin enables custom connectors, creates the Askable connector, then publishes it to the organisation. Each member then authenticates individually when they connect.
Part 1: Organisation Configuration
Sign in as an organisation admin.
First, make sure custom connectors are allowed at the organisation level:
From the file browser, click Admin, then Settings
Go to the AI section, then click Connectors in Figma Make
Toggle Allow custom connectors on
Then create and publish the Askable connector:
Open any Make file
In the chat box, click Add context, hover over Connectors, then select Manage
Go to the Created by you tab and click Create
Enter Name:
AskableEnter MCP server URL:
https://mcp.askable.com/mcpLeave Advanced settings empty. Askable uses OAuth 2.0, which is the default.
Click Create
Click Connect on the new connector and authenticate in Askable
Review the available Read and Write tools and enable the ones you want available. Write tools are disabled by default.
Click More, then select Publish
Review the details, tick the confirmation box, then click Publish
The connector is now available to everyone in your organisation under the tab with your organisation's name.
Part 2: Askable User Setup
Open any Make file
In the chat box, click Add context, hover over Connectors, then select Manage
Go to the tab with your organisation's name
Find Askable and click Connect
Authenticate in Askable
In chat, type @ and select Askable to scope a message to it. You can also prompt naturally, and Figma AI will call Askable when relevant.
Lovable
Setup happens in two stages. A workspace admin enables MCP server access at the workspace level. Each member then adds the Askable connector to their own account. Custom MCP connectors require a paid plan.
Part 1: Workspace Configuration
Sign in as a workspace admin or owner.
Go to Settings → Privacy & security
Enable MCP server access for the workspace
On Business and Enterprise plans, you can also approve individual MCP servers. Once a member has added Askable, you'll find it under Connectors → Chat connectors → Admin settings, where you can enable or disable it for everyone.
Part 2: Askable User Setup
In Lovable, open Connectors and go to Chat connectors
Click New MCP server
Server name:
AskableServer URL:
https://mcp.askable.com/mcpAuthentication: leave as OAuth, the default. Askable supports OAuth.
Click Add & authorize
Authenticate in Askable
Inside a project, click + in the prompt box, go to Connectors, and select Askable. You can also prompt naturally and Lovable will use Askable when relevant.
Tips for getting the best results
Be specific about what you want to know. The more context you give, the better the results. For example, "What concerns do first-time users in the UK have about the checkout flow?" will give you more targeted evidence than a broad question like "Tell me about checkout."
Use filters to narrow results. You can include filtering criteria directly in your questions to focus on the evidence that matters most. Available filters include:
Age range, e.g. "participants aged 18 to 35"
Location, e.g. "users in Australia" or "UK participants"
Gender, e.g. "female participants"
Time period, e.g. "from the last 3 months" or "studies run in January 2025"
Specific studies, e.g. "in the onboarding usability study"
Follow the evidence trail. Every response includes links back to the original highlights and recordings in Askable. Click through to verify quotes, watch the full session, or share the source with your team.
Share what you find. When you're ready to share, just ask your AI tool to create a report. It will generate an evidence-rich document in Askable with video highlights and reels embedded inline, then hand you a link. Your whole workflow stays in one place: explore in your AI tool, then share the link with your team.
Not sure what to ask? Once you're connected, you can ask your AI tool what you can do with the Askable MCP server and it will walk you through what's possible.
Try these example queries to get started:
"What are the biggest pain points in the checkout flow?"
"How do millennials talk about switching banks?"
"What studies do we have about onboarding?"
"Show me the transcript for that session"
"What are participants in the UK aged 25 to 34 saying about delivery fees?"
"What did female participants say about trust during checkout in the last 6 months?"
Frequently asked questions
Can I query across multiple teams?
Can I query across multiple teams?
Yes. If you have access to multiple teams or industry streams, the MCP server can switch context and search across them, bringing back a complete response considering all evidence returned.
What study types are supported?
What study types are supported?
Any study that captures transcript or recording data: moderated interviews, AI-moderated interviews, AI-moderated live website tests, unmoderated studies with recording enabled, and prototype tasks.
Is my data secure?
Is my data secure?
Every user authenticates individually, with no shared API keys. Access is scoped to your own workspace. Participant data follows Askable's existing privacy model. If someone leaves the team, their access is revoked without affecting anyone else.
What if an answer seems wrong?
What if an answer seems wrong?
Every response includes links to the source material. Click through to the original highlight clip or transcript to verify. Unlike a DIY tool, you're never stuck trusting a summary you can't check.
Can I create or modify studies through MCP?
Can I create or modify studies through MCP?
Not yet, but it's coming soon. The current version is read-only, so you can query and retrieve research data, but creating or editing studies still happens in the Askable app for now.
I'm having trouble connecting. What should I do?
I'm having trouble connecting. What should I do?
Make sure you're signed into your Askable account and that your account has access to research data. If the issue persists, contact Askable support.
