Sealgate MCP Server
Allows AI agents to read and send Discord messages through the Sealgate gateway, with policy enforcement and audit logging.
Allows AI agents to read and send iMessage messages through the Sealgate gateway, with policy enforcement and audit logging.
Allows AI agents to read and send Instagram DMs through the Sealgate gateway, with policy enforcement and audit logging.
Allows AI agents to read and send LINE messages through the Sealgate gateway, with policy enforcement and audit logging.
Allows AI agents to read and send Messenger messages through the Sealgate gateway, with policy enforcement and audit logging.
Allows AI agents to read and send Signal messages through the Sealgate gateway, with policy enforcement and audit logging.
Allows AI agents to read and send Telegram messages through the Sealgate gateway, with policy enforcement and audit logging.
Allows AI agents to read and send WhatsApp messages through the Sealgate gateway, with policy enforcement and audit logging.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Sealgate MCP ServerReview recent agent sessions and audit events for blocked actions."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Sealgate MCP Server
The Model Context Protocol (MCP) server for Sealgate, the AI data-leak-prevention platform: a security gateway and data firewall that sits between AI agents (Claude, ChatGPT, Cursor, Copilot) and your organisation's data and tools.
This server is a thin MCP proxy. It forwards a small set of tool calls to a
Sealgate gateway, which you configure with two environment variables.
Sealgate runs managed release hosts (the remote MCP gateway at mcp.sealgate.ai
and the Management API at dashboard.sealgate.ai), and demo or self-hosted orgs
run their own, so this bridge stays host-agnostic: you supply the URL and key.
Tools
Tool | What it does |
| List the MCP servers governed by your Sealgate gateway, with access-control classification and connection status. |
| Review recent agent sessions and audit events: what agents did, which data flowed, and any blocked actions. |
Related MCP server: SINT Protocol
Connect to the hosted gateway
Sealgate runs a managed remote MCP gateway at https://mcp.sealgate.ai/mcp. It
is a per-user proxy that aggregates every MCP server you have enabled behind one
Streamable HTTP endpoint and enforces Sealgate's access-control policies on every
call. There is nothing to install: point your client at the endpoint and sign in
through the browser.
Endpoint: https://mcp.sealgate.ai/mcp
Transport: streamable HTTP (remote, not stdio)
Auth: OAuth 2.1 in the browser, no key to pasteClient | Add it |
Claude | Add to Claude opens the Add custom connector dialog with the name and URL filled in. Claude flags it as suggested by an external link; that is expected. On Team and Enterprise plans an admin adds it. |
ChatGPT | Settings → Connectors → Advanced settings → turn on Developer mode. Back on Connectors, click Create, paste the endpoint, and name it. Start a new chat so the tools menu refreshes. |
Claude Code |
|
Cursor | Not working? Add the endpoint by hand under Settings → MCP. |
VS Code | Add to VS Code requires Copilot agent mode. |
Goose | Adds it as an extension over streamable HTTP. |
Grokbot | Add a remote MCP server pointing at |
Any MCP client | Add a remote server at |
Every one-click button routes through an https:// install URL, since GitHub
strips custom URL schemes such as cursor:// from links.
How OAuth works
The gateway is an OAuth 2.1 authorization server, so most clients connect with no
API key at all. The flow uses dynamic client registration (RFC 7591) and client
ID metadata documents, mandatory PKCE (S256), and issues refresh tokens via the
offline_access scope, so a session stays connected without re-authenticating
every hour. Discovery, consent, and token endpoints all live on the gateway
origin, so self-hosted single-origin deployments work with no extra
configuration.
API key in the URL
Clients that cannot run an OAuth flow can pass a Sealgate API key as a URL path segment instead:
https://mcp.sealgate.ai/mcp/{api_key}/?client={label}Replace {api_key} with the key from your dashboard and {label} with an
optional session label. This is not OAuth and not an Authorization header; the
key travels in the path.
mcp.sealgate.ai is the managed release host. Demo and self-hosted orgs run
their own gateway host, so substitute your own URL where needed.
Connect your messaging apps
The same gateway fronts Sealgate's messaging connectors (the Beeper desktop app
plus the sealgate-stdiod tunnel), so any client above can read and send across
your chat networks. Every message an agent reads or sends passes through the
same policy checks and lands in the same audit log as every other tool call.
Network | Guide |
Telegram | |
iMessage | |
Signal | |
LinkedIn DMs | |
Messenger | |
Discord | |
X | |
LINE | |
Per-client guides | Codex + iMessage, Claude + LinkedIn, ChatGPT + WhatsApp, and all 100 at https://sealgate.ai/connect |
Codex + iMessage in one command: the companion repo Edison-Watch/codex-imessage wires Codex to iMessage through the gateway with a single install step.
Configuration
Set two environment variables, both issued or hosted by your organisation:
Variable | Description |
| Base URL of your Sealgate Management API (e.g. |
| Sealgate API key from your dashboard. |
If either is unset, every tool returns a clear configuration message instead of
failing, so registry probes and --help never crash.
Install
Prefer the hosted gateway above. The npm package is the stdio alternative for
clients that cannot use a remote server; it runs via npx and needs the two
environment variables from Configuration. Also published in the official MCP
Registry as ai.sealgate/gateway and io.github.Edison-Watch/sealgate-mcp.
{
"mcpServers": {
"sealgate": {
"command": "npx",
"args": ["-y", "@sealgate/mcp"],
"env": {
"SEALGATE_GATEWAY_URL": "https://dashboard.sealgate.ai",
"SEALGATE_API_KEY": "your-sealgate-api-key"
}
}
}
}VS Code spells the same block as servers with "type": "stdio". Ready-made
files for each client live in examples/.
Usage
# Run over stdio (default transport)
npx -y @sealgate/mcp
# Show usage
npx -y @sealgate/mcp --helpSet SEALGATE_MCP_TRANSPORT=http (with an optional PORT, default 3000) to
serve streamable HTTP instead of stdio.
Develop
Requires Bun.
bun install
bun run src/index.ts --help # run from source
bun test # run tests
bun run build # bundle to dist/
make ci # lint, typecheck, dead-code, and the restLinks
Website: https://sealgate.ai
Docs: https://docs.sealgate.ai
Developer docs, API reference, and integration guides: SealGate developer docs
Connect guides: https://sealgate.ai/connect
Contact: hello@sealgate.ai
License
MIT. Copyright GPU-EVM LTD (Sealgate). See LICENSE.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
No tool schema history has been recorded yet.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Connectors
Security & DLP proxy for MCP: tool-poisoning scans, PII redaction on tool args/results. Beta.
Zero-secret MCP gateway for AI agents: risk-scored, audited calls with human-in-the-loop approval.
Search, vet & assemble MCP servers from your agent: verified tools, risk labels, and trust scores.
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
Related MCP Servers
AlicenseNot gradedqualityAmaintenanceOpen-source MCP proxy that enforces security policies, content scanning, and audit logging between AI agents and tool servers25AGPL 3.0- AlicenseAqualityAmaintenanceSecurity-enforcing MCP proxy that sits between an AI agent and any number of downstream MCP servers, intercepting every tool call through a capability-token policy gateway that can allow, deny, or escalate to human approval before the call reaches any real tool. It also exposes built-in operator tools for approval workflows, audit trail queries, token management, voice/HUD output, and hierarchical2112Apache 2.0
- AlicenseNot gradedqualityAmaintenanceSecurity gateway for MCP tool calls. Sits between your LLM client and MCP servers, enforcing per-tool policies (allow/block/approve/read-only), logging every call, and pausing dangerous operations for human approval in terminal or Slack.41MIT

evav-gatewayofficial
AlicenseNot gradedqualityBmaintenanceGoverned MCP gateway that lets AI agents call tools with policy enforcement, prompt-injection screening, a kill-switch, and tamper-evident signed audit logs.Apache 2.0
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/Edison-Watch/sealgate-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server