The server itself exposes 30 tools, and some of them can write to your Jam account. Behind TrustGate an agent only sees the tools you allow, requests and responses are inspected, and every call is logged. Anything outside the allowlist is denied before it reaches Jam.

How is this different from connecting Jam MCP directly?

Connected directly, an agent sees every tool the server exposes and holds the credential itself. Through TrustGate you choose which tools each agent gets, and every call is logged against the agent that made it.

What if the Jam MCP API changes?

We track the Jam MCP schema, so the tool catalogue stays current. Tools you have already allowed keep working, and anything new stays denied until you enable it.

Can I give different agents different Jam tools?

Yes. Access is set per agent: one can get a handful of read tools, another the full set. You grant the whole MCP or single tools within it.

Where can I see what my agents did in Jam?

Every call is logged against the agent that made it, with the tool it used. Activity can be sent on to your own monitoring.

Do I need to run the Jam MCP server myself?

No. TrustGate connects to the Jam MCP server and sits in front of it. Your agent only talks to your TrustGate URL, never to Jam directly.

Connect Jam MCP to your AI agents, securely.

One secured endpoint for every Jam call your AI agents make, scoped per user, inspected, and logged.

Get started for free

Get a demo

Try Jam MCP now

What your agents can ask once Jam MCP is connected.

Select the tools your AI agents can call

30 tools from the Jam MCP schema.

Showing 30 tools

getDetailsreadReturns a quick snapshot of the Jam: who made it, what happened, and which other tools to try next.
getConsoleLogsreadReturns the console logs from the Jam, with logLevel, limit and page-cursor filters to reduce noise.
getNetworkRequestsreadLists every web request from the Jam as JSON, filterable by status code, content type, host and method. The bodies parameter defaults to errors but accepts all, which returns full request and response bodies from the recorded session.
getUserEventsreadReturns each click, input and page navigation from the recording in plain language.
getMetadatareadReturns custom key-value metadata set through the SDK: user IDs, app versions, feature flags and more.
getScreenshotsreadReturns every screenshot from a screenshot Jam for visual inspection. Not available for Instant Replay Jams.
getVideoTranscriptreadReturns the spoken transcript from a video Jam where the microphone was enabled, in WebVTT format with timestamps.
getVideoChaptersreadReturns AI-generated chapter markers for a video Jam: titled segments with start and end timestamps, text only.
getFramesreadReturns still frames showing what was on screen, at specific timestamps, evenly across a window, or as one labeled overview grid spanning the recording. It works only on video Jams hosted on Cloudflare Stream.
searchJamsreadSearches Jams by text across titles, descriptions, comments, linked issues and transcripts, with filters for author, folder, type, date, page URL, source, linked issue, environment and exclusions.
listMembersreadBrowses the team members in the workspace.
listFoldersreadBrowses the available folders.
createCommentwriteAdds a Markdown comment to a Jam, optionally mentioning a teammate to notify them.
editCommentwriteRewrites a comment left earlier, so an agent can correct its own note instead of stacking a follow-up.
deleteCommentwriteDeletes a comment you wrote, along with its attachments.
addReactionwriteReacts to a comment, useful for acknowledging one already acted on.
removeReactionwriteRemoves a reaction left on a comment.
updateJamwriteRenames a Jam, rewrites its description, or moves it between folders by folder ID or name.
createFolderwriteCreates a folder to organize Jams into, returning its ID and short ID for a follow-up updateJam.
updateFolderwriteRenames a folder, addressed by its name, ID or short ID.
deleteJamhigh riskDeletes a Jam. It leaves lists and search, and there is no way to restore it.
deleteFolderhigh riskDeletes a folder and every Jam inside it, none of them restorable. It returns how many Jams went with it.
listRecordingLinksreadLists the team's recording links, newest first, with cursor pagination and an includeRevoked option.
getRecordingLinkreadReturns one recording link by its public ID.
listRecordingLinkJamsreadLists the Jams recorded through a link, by public ID, with cursor pagination.
listRecordingUrlsreadLists the team's connected recording domains.
createRecordingLinkhigh riskCreates a reusable recording link, optionally targeting a folder, connected domain or expiration. Anyone holding the link can record into the workspace.
updateRecordingLinkwriteEdits a link's name, folder, reference, Jam title, expiration or metadata by public ID. Revoked links cannot be edited.
deleteRecordingLinkwriteRevokes a link by its public ID. This is a soft delete, so collected Jams stay.
getRecordingUrlVerifyLinkwriteStarts verifying a connected domain and returns a link a human opens to confirm it.

What Jam MCP exposes

Jam MCP groups its 30 tools into 5 categories, each with a suggested starting policy.

All 30 tools as JSON, ready to paste.

Category	Covers	Access	Suggested policy
Jam Data	A snapshot of the Jam and who made it, console logs with level and paging filters, every network request as JSON with optional bodies, user clicks, inputs and navigations in plain language, custom key-value metadata from the SDK, screenshots, the spoken transcript in WebVTT, AI-generated chapter markers, structured video analysis, and still frames at chosen timestamps or as an overview grid	Read	Read-only but the richest surface here. getNetworkRequests with bodies set to all returns full request and response bodies from a real session; getFrames and getScreenshots return the screen as it was. Treat the output as production data, not as logs
Recording Links	Listing the team's recording links with cursor pagination and a revoked option, reading one by public ID, listing the Jams recorded through it, creating a reusable link optionally targeting a folder, connected domain or expiration, editing a link, revoking it, listing connected recording domains and starting domain verification	Read / Write	createRecordingLink produces an externally usable link that writes new Jams into the workspace, which makes it the one outward-facing action on this connection. deleteRecordingLink is a soft delete, so collected Jams stay
Comments & Reactions	Adding Markdown comments to a Jam with optional teammate mentions that notify them, rewriting a comment left earlier, deleting one you wrote along with its attachments, and adding or removing a reaction on a comment	Write	createComment can mention a teammate, which sends them a notification, so this is the agent speaking to people rather than to a record. editComment exists so an agent corrects itself instead of stacking follow-ups
Organization	Renaming a Jam, rewriting its description or moving it between folders, deleting a Jam, creating a folder and getting back its IDs, renaming a folder by name or ID, and deleting a folder together with every Jam inside it	Read / Write	deleteFolder takes every Jam inside with it and Jam states plainly that none of them are restorable; it returns how many went. deleteJam is equally final for a single recording
Search & Browse	Searching Jams by text across titles, descriptions, comments, linked issues and transcripts with filters for author, folder, type, date, page URL, source, linked issue key, environment and exclusions, plus browsing team members and folders	Read	Read-only, but searchJams reaches across transcripts and comments, so a broad query can surface recordings from parts of the workspace the request was not about

Connect Jam MCP to your AI agent

Four steps, under ten minutes.

01
Sign in to NeuralTrust
Open the NeuralTrust platform, sign in and go to Getting started, then Connect your tools.
02
Choose your agent
When asked which agent you use, pick Claude Code, Codex, Cursor, Windsurf, Copilot or your own code.
03
Connect Jam
Find Jam in the tool list and connect it.
04
Connect your agent
Follow the steps on screen to add the config to your agent, then send the test prompt.

Permissions down to the single tool

You do not allow Jam MCP or block it. You allow the exact tools each agent can call, one by one.

Jam MCP	30 TOOLS AVAILABLE
Code review agent	4 tools, read only
Issue triage bot	9 tools
Platform team assistant	all 30 tools

Tool by toolConnecting the server is not the unit of access. Each tool is granted on its own.
Default denyA tool you have not enabled is never exposed to the agent.
Approval on writesTools that change or delete things can require a human to confirm.

Is Jam MCP secure?

One endpoint instead of raw credentials. Every call is checked against policy and logged.

AI agentsConnect through your MCP client configuration.	NeuralTrustApply controls based on your configuration.	Jam MCP30 documented tools at mcp.jam.dev/mcp over OAuth. MCP mirrors existing Jam permissions and grants no access a user could not already get in the web or mobile apps; connected tools can be revoked individually under Settings. Some tools use Google's Gemini, with Jam opting out of training on customer data.
Coding assistants	Connection & credentialsper user	Jam Data9 tools
Internal copilots	Tool access policiesdeny by default	Recording Links8 tools
Workflow automations	Request & response inspectioninline	Comments & Reactions5 tools
Compatible MCP clients	Activity monitoringalerts and logs	All other tools8 tools