How is this different from connecting Egnyte 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 Egnyte MCP API changes?

We track the Egnyte 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 Egnyte 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 Egnyte?

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 Egnyte MCP server myself?

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

Connect Egnyte MCP to your AI agents, securely.

Q: Is Egnyte MCP secure?

The server itself exposes 28 tools, and some of them can write to your Egnyte 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 Egnyte.

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

Get started for free

Get a demo

Try Egnyte MCP now

What your agents can ask once Egnyte MCP is connected.

Select the tools your AI agents can call

28 tools from the Egnyte MCP schema.

Showing 28 tools

searchreadSearches files by full-text query, returning ranked documents with content snippets. The index covers file names, document content across all supported file types, and metadata.
advanced_searchreadAdvanced search with custom metadata, date ranges for modification and upload, folder paths, file and folder query fields, namespaces, sorting and similarity search.
fetchreadFetches the content of a specific document by its identifier. Content is truncated to 10,000 characters.
ask_documentreadAsks a question about a specific document using Egnyte AI, with optional citations.
summarize_documentreadGenerates an AI-powered summary of a document.
get_ai_assistant_statusreadChecks the status of an ask_ai_assistant request. On completion it returns the response text and citations.
ask_ai_assistantwriteSubmits a question to the Egnyte AI Assistant and returns an execution ID to poll. It can be scoped to up to 20 files and folders combined, or set to search all content the user can access; the two modes cannot be combined.
list_agentsreadLists the active Agents the user created or that are shared with them, with each Agent's ID, name, description and category. Egnyte's built-in Agents are not included.
get_agent_ask_statusreadChecks the status of an ask_agent request. Responses can take up to two minutes, and AWAITING_USER_CONFIRMATION is a final state for MCP clients.
ask_agentwriteSubmits a question of up to 35,000 characters to an Agent and returns IDs to poll. Optional folder and file IDs add content to the Agent's configured scope for that question only.
ask_knowledge_basereadQueries a specific Knowledge Base, with optional citations. Deprecated in favour of ask_agent.
list_knowledge_basesreadLists the available Knowledge Bases with filtering and sorting. Deprecated in favour of list_agents.
list_filesystem_by_pathreadLists files and folders by path, with options to include permissions, custom metadata and statistics.
get_file_contentreadReturns the text content of a file as markdown or plain text, with pagination. The file is identified by group ID or path.
get_download_urlreadGenerates a single-use signed URL for direct binary download, for when the raw stream is needed such as PDFs, images and archives.
list_metadata_namespacesreadReturns all available metadata namespaces and their field definitions for the domain. It is the prerequisite for set_file_metadata.
list_projectsreadLists the project folders in the Egnyte domain, optionally filtered by folder path.
create_folderwriteCreates a new folder. It fails if the folder already exists or the parent folders do not.
upload_filewriteUploads a UTF-8 text file, creating it or adding a new version if one already exists at the path. Maximum size is 8 MB.
get_upload_urlwriteGenerates a single-use signed URL for direct binary upload of any Egnyte-compatible file type, creating a new file or a new version. Transfers above 200 MB may not complete reliably.
set_file_metadatahigh riskSets custom metadata on a file. It replaces all existing metadata in the specified namespace rather than merging or appending.
list_commentsreadLists comments for a specific file or all comments in the domain, filterable by time range.
get_commentreadReturns detailed information for a specific comment by its identifier.
list_linksreadLists Egnyte share links with filtering by type, accessibility, path, username and creation date.
get_link_detailsreadReturns detailed information about a specific Egnyte link.
create_commentwriteCreates a comment, or note, on an existing file.
create_linkhigh riskCreates a share link for a file or folder and returns its URL. It sets type, accessibility, expiry date, password, recipients, protection and whether to send an email and notify, so it can make internal content externally reachable and mail the link out.
whoamireadReturns the Egnyte identity this connector is authenticated as: the username and domain.

What Egnyte MCP exposes

Egnyte MCP groups its 28 tools into 7 categories, each with a suggested starting policy.

All 28 tools as JSON, ready to paste.

Category	Covers	Access	Suggested policy
File System & Projects	Listing files and folders by path with metadata and permissions, reading text content with pagination, creating folders, uploading text files, generating signed URLs for direct binary upload and download, setting custom metadata, discovering metadata namespaces, and listing project folders	Read / Write	set_file_metadata replaces every value in a namespace rather than merging, so read list_metadata_namespaces first. The signed-URL pair moves bytes outside the server, so log the call even though the transfer is not visible to it
Collaboration	Listing and reading comments on a file or across the domain, creating a comment, listing and inspecting Egnyte links with their accessibility and expiry, and creating a share link for a file or folder	Read / Write	create_link is the one externally visible action on this connection: it sets accessibility, can email recipients directly and optionally applies a password and expiry. Require approval on every call and log the resulting URL
AI-Powered	Asking questions of a specific document with optional citations, generating an AI summary, submitting a question to the Egnyte AI Assistant scoped to files and folders or to everything the user can reach, and polling for its answer	Read / Write	ask_ai_assistant is asynchronous and can be scoped to up to 20 files and folders or set to search all Egnyte content; the two modes cannot be combined. Prefer the scoped form so the question does not range across the whole domain
Search & Discovery	Full-text search returning ranked documents with content snippets across file names, document content and metadata, advanced search with custom metadata, date ranges, folder paths and similarity, and fetching a document's content by its identifier	Read	Read-only and bounded by the user's own permissions. fetch truncates content to 10,000 characters, so a full read needs get_file_content instead
Agents	Listing the active Egnyte Agents the user created or that are shared with them, submitting a question to one with optional extra files and folders for that question only, and polling for the answer	Read / Write	Agents are limited to Power Users and Admins. A request can end in AWAITING_USER_CONFIRMATION, a deliberate dead end meaning the Agent needs a confirmation or external sign-in that cannot be given through MCP
Knowledge Base	Querying a specific Knowledge Base and listing the available ones with filtering and sorting	Read	Deprecated. Knowledge Bases are being converted to Agents through October 2026, with IDs preserved, and these two return an error once a domain is migrated
Identity	The Egnyte username and domain this connector is authenticated as	Read	Low risk and worth calling first: everything on this connection is bounded by that user's permissions, so the identity is the scope

Connect Egnyte 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 Egnyte
Find Egnyte 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 Egnyte MCP or block it. You allow the exact tools each agent can call, one by one.

Egnyte MCP	28 TOOLS AVAILABLE
Code review agent	4 tools, read only
Issue triage bot	9 tools
Platform team assistant	all 28 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 Egnyte 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.	Egnyte MCP28 documented tools at mcp-server.egnyte.com/mcp over OAuth 2.0, each labelled Read or Write by Egnyte. Gen 3 plans without the AI Add-On get six tools only. Access is limited to what the authenticated user can already reach, and Egnyte states the server does not store or cache accessed content.
Coding assistants	Connection & credentialsper user	File System & Projects9 tools
Internal copilots	Tool access policiesdeny by default	Collaboration6 tools
Workflow automations	Request & response inspectioninline	AI-Powered4 tools
Compatible MCP clients	Activity monitoringalerts and logs	All other tools9 tools