MCP - Jam Documenation

AI agent debugging a Jam via MCP, with calls to getConsoleLogs, getNetworkErrors, and getUserEvents tools

Jam MCP (Model Context Protocol) lets your AI coding tools open a Jam recording and automatically load its full context (video, user events, console logs, errors, and network requests) right where you’re working. Paste a Jam link into any supported MCP client and the context flows directly to your agent. Jam captures console logs, network requests, user events, and your chosen metadata automatically when you record. With MCP, that context flows into Claude Code, Cursor, VS Code, and Windsurf without copy-pasting.

Prerequisites

You need a Jam recording before your agent can use MCP.

Install the Jam Chrome extension

Install the Chrome extension from the Chrome Web Store.

Record a Jam

Hit Record to capture your screen and voice.

Copy your Jam link

Stop the recording. Your Jam link is ready for an agent to consume via MCP.

Step 1: Configure your MCP client

Open settings

From the ChatGPT home page, go to Settings.

Navigate to connectors

Click Apps & Connectors.

Connect Jam

Select Jam and click the Connect button.

Authorize

Install Claude Code in your IDE, then run the following command in your terminal:

claude mcp add Jam https://mcp.jam.dev/mcp -t http -s user

Open or create the config file at ~/.codex/config.toml
Add the following to your config file and save:

[mcp_servers.jam]
url = "https://mcp.jam.dev/mcp"

Add the following to your Cursor MCP configuration:

{
  "mcpServers": {
    "jam": {
      "url": "https://mcp.jam.dev/mcp"
    }
  }
}

Restart Cursor after making this change for it to take effect.

Open the command palette

Press Ctrl/Cmd + P and enter >MCP: Add Server.

Select the transport type

Select HTTP (HTTP or Server-Sent Events).

Enter the server URL

Enter https://mcp.jam.dev/mcp.

Name the server

Enter Jam.

Your mcp.json file should look like this:

{
  "servers": {
    "Jam": {
      "url": "https://mcp.jam.dev/mcp",
      "type": "http"
    }
  }
}

Open the command palette

Press Ctrl/Cmd + P and enter >MCP: Add Server.

Select the transport type

Select HTTP (Server-Sent Events).

Enter the server URL

Enter https://mcp.jam.dev/mcp.

Name the server

Enter Jam.

Your mcp.json file should look like this:

{
  "mcp": {
    "servers": {
      "Jam": {
        "type": "http",
        "url": "https://mcp.jam.dev/mcp"
      }
    }
  }
}

Edit your opencode.json:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "jam": {
      "type": "remote",
      "url": "https://mcp.jam.dev/mcp"
    }
  }
}

Then run the following command to start the OAuth2 authorization flow:

opencode mcp auth jam

Step 2: Start using Jam MCP

Once your MCP client is configured:

Copy your Jam link

Copy one or more Jam links from your workspace.

Prompt your agent

Paste the Jam link into your agent’s chat and describe what you need. See the prompting tips and examples below.

Authenticate

Authenticate your IDE with Jam when prompted.

Approve tool requests

Approve any tool requests the agent makes as it reads your Jam data.

Available MCP tools

Tool	Description
`getDetails`	Get a quick snapshot of the Jam: who made it, what happened, and which other tools to try next.
`getConsoleLogs`	Grab the console logs from the Jam. Supports `logLevel` and `limit` filters to reduce noise.
`getNetworkRequests`	List every web request from the Jam as JSON. Supports `statusCode`, `contentType`, `host`, and `limit` filters.
`getScreenshot`	Get every screenshot from a screenshot Jam for visual inspection.
`getUserEvents`	Read each click, input, and page navigation change in plain language.
`getMetadata`	Read custom key-value metadata set via the `jam.metadata()` SDK: user IDs, app versions, feature flags, and more.
`getVideoTranscript`	Get the spoken transcript from a video Jam where the microphone was enabled, in WebVTT format with timestamps.
`analyzeVideo`	Analyze a Jam video recording to extract insights, detect issues, and provide structured feedback.
`getFrames`	See what was on screen in a video Jam. Return frames at specific timestamps, evenly across a window, or one labeled overview grid spanning the whole recording.
`listJams`	Search and filter Jams by text, type, folder, author, URL, or date.
`listMembers`	Browse your team members.
`listFolders`	Browse available folders.
`createComment`	Add Markdown comments to a Jam.
`updateJam`	Move Jams between folders using a folder ID or folder name.
`listRecordingUrls`	List the team’s connected recording domains.
`getRecordingUrlVerifyLink`	Start verifying a connected domain and return a link a human opens to confirm it.
`listRecordingLinks`	List the team’s recording links, newest first, with cursor pagination and an `includeRevoked` option.
`getRecordingLink`	Get one recording link by its public ID.
`listRecordingLinkJams`	List the Jams recorded through a link, by public ID, with cursor pagination.
`createRecordingLink`	Create a reusable recording link, optionally targeting a folder, connected domain, or expiration.
`updateRecordingLink`	Edit a link’s name, folder, reference, Jam title, expiration, or metadata by public ID (revoked links can’t be edited).
`revokeRecordingLink`	Revoke a link by its public ID (soft-delete, so collected Jams stay).

getScreenshot and analyzeVideo are not available for Instant Replay Jams. All other MCP tools work and return data for Instant Replay Jams.

Seeing what was on screen

getVideoTranscript and analyzeVideo describe a video Jam in text, but sometimes your agent needs the actual pixels. getFrames returns still frames from the recording as images, so your agent can look at the screen at the moments that matter. Call it three ways:

Overview grid. Set overview: true to get one image: frames evenly spaced across the whole video, stitched into a labeled grid where each cell shows its timestamp. The frame count scales with the video’s length (6 for short clips up to 16 for long ones). This is the fastest way for an agent to orient itself on a visual bug before drilling in.
Specific timestamps. Pass at with one or more times in milliseconds to capture those exact moments.
Window. Pass fromMs, toMs, and count to sample evenly-spaced frames across a range.

Use size (small, medium, large; default medium) to trade detail against your agent’s context budget. Each frame comes back as an image preceded by a t=<ms> text anchor so the agent knows which moment it’s looking at.

getFrames works only on video Jams whose recording is hosted on Cloudflare Stream. Screenshot Jams, and videos not yet on Cloudflare, return a short text block explaining why instead of images.

Prompting tips

Use Jam MCP to help with bug analysis, product feedback, and giving agents the context they need to debug issues.
Be as specific as you would with another engineer for the best results.
Start with small features or bug fixes. Tell your agent to propose a plan using Jam MCP and verify its plan before moving on to suggested edits.

Examples

Here are three ways the team at Jam uses Jam MCP in their own workflow.

Bug analysis

Tomasz (Product Engineer) uses Jam MCP to debug issues by providing bug context to Cursor to ship a fix.

Review this Jam [Jam Link] and analyze the problem, cross-reference it with the existing codebase, and prepare a detailed plan for implementation.

Stack: Jam recordings + Jam MCP + Cursor

Product feedback triage

Frits (Product Manager) lets Claude review and analyze customer Jams to spot common issues and patterns, then turns findings into grouped Linear tickets.

Analyze these Jams, perform root-cause analysis, and organize them into efficient work packages to minimize engineers' context switching; create Linear tickets from the analysis.

[Jam Link]

Stack: Jam customer recordings + Jam MCP + Claude Desktop + Linear MCP

Implementation planning from design feedback

Martin (Product Designer) records product and design feedback with the Jam extension. Jam MCP provides the context for Claude Code to plan and ship a ready-to-review PR.

I recorded product feedback in a Jam. Create an implementation plan for the changes requested in this Jam:

[Jam Link]

Stack: Jam extension + Jam MCP + Claude Code inside Cursor

Managing MCP access

Go to Settings → MCP to see connected tools. Select any tool to revoke or disconnect access at any time.

MCP mirrors your existing Jam permissions. It does not grant new access beyond what a user can already see in the Jam web or mobile apps. Use your normal admin controls for members and roles, or SSO in Settings.

FAQ

Can I use MCP without the Chrome extension?

Yes. You only need an existing Jam link. Anyone (you, your workspace, or a customer) can record a Jam. If you want to record and create new Jams yourself, install the Chrome extension or use the iOS app.

Can I paste multiple Jam links at once?

Yes. You can paste more than one link, but we recommend one Jam at a time so your agent doesn’t hit context window limits.

Does Jam MCP send my Jam data to third parties?

Some MCP tools use Google’s Gemini. Jam opts out of training on customer data and takes steps to de-identify it. See the Security page for more details on Jam’s AI data policy.

Does MCP work the same with Instant Replay Jams?

Not exactly. getScreenshot and analyzeVideo are not available for Instant Replay Jams. All other MCP tools still work and return data and context.

How can admins control what MCP can access?

MCP mirrors your existing Jam permissions. It doesn’t grant new access. Use your normal admin controls for members and roles, or SSO in Settings. Nothing is exposed via MCP that a user couldn’t already see in the Jam web or mobile apps.

​Prerequisites

​Step 1: Configure your MCP client

​Step 2: Start using Jam MCP

​Available MCP tools

​Seeing what was on screen

​Prompting tips

​Examples

​Bug analysis

​Product feedback triage

​Implementation planning from design feedback

​Managing MCP access

​FAQ

Prerequisites

Step 1: Configure your MCP client

Step 2: Start using Jam MCP

Available MCP tools

Seeing what was on screen

Prompting tips

Examples

Bug analysis

Product feedback triage

Implementation planning from design feedback

Managing MCP access

FAQ