Prototype · Preview: every server in the official registry, synced every 30 minutes. Sign in with Google to write reviews.Prototype · updated every 30 min
PopNx MCPMCP servers for people who use AI · explained in plain words

Unified AI System MCP Gateway

by happy520aiversion 0.8.0i★ 10 (Stars for the whole repo)iRuns on your computerOfficial site ↗
Registry nameio.github.happy520ai/unified-ai-system

How to install

What it does

Self-hosted MCP gateway for Codex, Cursor, and Cline with provider-free prompt enhancement.

  • Connects Runs on your computer
  • Account or key No key needed
  • Category Search & AI memory

Use together with

Skills that teach the AI this kind of work; they work alongside this one

timesfm

TimesFM (Time Series Foundation Model) is a pretrained time-series foundation model developed by Google Research for time-series forecasting.

by google-research★ 35,6381 skills

Skill · How-to guide

Shadowbroker

Open-source intelligence for the global theater. Track everything from the corporate/private jets of the wealthy, and spy satellites, to seismic events in one unified interface. Hook an AI agent up to have it parse through data and find previously unseen correlations. The knowledge is available to…

by BigBodyCobain★ 11,6051 skills

Skill · How-to guide

google-maps-scraper

scrape data from Google Maps. Extracts data such as the name, address, phone number, website URL, rating, reviews number, latitude and longitude, reviews,email and more for each place

by gosom★ 8,2361 skills

Skill · How-to guide

Why it ranks here

Lists are ordered by a usage score (out of 100) built from these parts

Popularity6/80GitHub stars or weekly downloads: the larger counts 60, the smaller 20
Freshness12/12Full marks for a release or code change within 30 days
Still answering0/8The cloud endpoint answered at the last check
Usage score18/100

Quality: Passed 3/3 basic checks (shown, not used for ordering) · Momentum: No data yet (Used only on the Trending tab)

Install

Pick the AI app you use, then copy the settings.

Let your AI install it with one sentence

Please install Unified AI System MCP Gateway. Follow https://mcp.popnx.com/i/happy520ai-unified-ai-system.md, show me the commands and wait for my OK before running them.

Paste it into an AI that can work on your computer (such as Claude Code or Cursor). It reads our steps, shows you the commands and waits for your OK. No key, no sign-up. · See the full text your AI reads

Passed 3/3 checks · Runs on your computer · ✓ Commands checked Nov 14, 2026

First: which AI do you use?

Pick your app and the steps appear below.

Works with Claude, with some setup first

Install Docker on your computer first (one time) · Only the Claude desktop app (Windows or Mac), not the web or phone

Let AI install it

Copy this message into a Claude chat. The AI checks your computer, tells you which file it will change, and waits for your OK every time.

Please install the MCP server Unified AI System MCP Gateway for Claude on this computer, in this order: 1) Check Docker with docker --version. If it is missing, tell me to install it from https://www.docker.com/products/docker-desktop/; do not install it yourself. 2) Use only this config (already checked by PopNx MCP): { "mcpServers": { "unified-ai-system": { "command": "docker", "args": [ "run", "-i", "--rm", "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0" ] } } } 3) Tell me the full path of the file you will change, or the command you will run, and wait for my OK. 4) No key is needed. 5) Do what we agreed, then check that the Unified AI System MCP Gateway tools really show up.

This message has no key in it. If a key is needed, the AI will ask you to paste it yourself into the file or the app’s field.

  1. Install Docker (if you don’t have it)

    Already installed? Check first · Open Terminal, type this command and press Enter

    docker --version
    • A version number such as Docker version … = installed, skip this step
    • "command not found" (on Windows: "is not recognized") = not installed yet, or installed but Terminal needs to be closed and reopened
    How to open Terminal

    Mac: press ⌘ + Space, type Terminal, press Enter

    Windows: open Start, type Windows PowerShell, then Open

    Not installed: Install Docker Desktop: on Mac open Docker.dmg and drag the icon to Applications (pick Apple silicon or Intel to match your Mac) · on Windows double-click Docker Desktop Installer.exe · then open the Docker Desktop app and leave it running. No curl command needed

    Open the Docker download page ↗

    When it’s done, close Terminal, open it again and run the same command. You should see a version number: docker --version

    Install once; it works for everything that needs Docker

  2. Open the Claude desktop app’s settings file

    SettingsDeveloperEdit Config

  3. Paste these settings and save the file

    {
      "mcpServers": {
        "unified-ai-system": {
          "command": "docker",
          "args": [
            "run",
            "-i",
            "--rm",
            "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0"
          ]
        }
      }
    }
  4. Quit Claude and open it again

  5. Try it

    Use Unified AI System MCP Gateway and tell me what you can do for me, then show me one thing

    ✓ If Claude says it is using Unified AI System MCP Gateway or shows a Unified AI System MCP Gateway tool in the answer, you are connected

Menu names checked against Anthropic’s help page on Nov 11, 2026 · your screen may differ slightly by app version

ChatGPT cannot connect Unified AI System MCP Gateway yet

Unified AI System MCP Gateway runs on your computer, but ChatGPT can only connect servers that run on the web

What you can do instead

A · Use an app that can connect Unified AI System MCP Gateway

  • Install Docker on your computer first (one time) · Only the Claude desktop app (Windows or Mac), not the web or phone
  • Install Docker on your computer first (one time)
  • Install Docker on your computer first (one time)
  • Install Docker on your computer first (one time)

B · Others in Search & AI memory that work with ChatGPT

Picked from the most used in the same category, only ones that connect by address with no key

Works with Claude Code, with some setup first

Install Docker on your computer first (one time)

Let AI install it

Copy this message into a Claude Code chat. The AI checks your computer, tells you which file it will change, and waits for your OK every time.

Please install the MCP server Unified AI System MCP Gateway for Claude Code on this computer, in this order: 1) Check Docker with docker --version. If it is missing, tell me to install it from https://www.docker.com/products/docker-desktop/; do not install it yourself. 2) Use only this config (already checked by PopNx MCP): claude mcp add --scope user unified-ai-system -- docker run -i --rm ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0 3) Tell me the full path of the file you will change, or the command you will run, and wait for my OK. 4) No key is needed. 5) Do what we agreed, then check that the Unified AI System MCP Gateway tools really show up.

This message has no key in it. If a key is needed, the AI will ask you to paste it yourself into the file or the app’s field.

  1. Install Docker (if you don’t have it)

    Already installed? Check first · Open Terminal, type this command and press Enter

    docker --version
    • A version number such as Docker version … = installed, skip this step
    • "command not found" (on Windows: "is not recognized") = not installed yet, or installed but Terminal needs to be closed and reopened
    How to open Terminal

    Mac: press ⌘ + Space, type Terminal, press Enter

    Windows: open Start, type Windows PowerShell, then Open

    Not installed: Install Docker Desktop: on Mac open Docker.dmg and drag the icon to Applications (pick Apple silicon or Intel to match your Mac) · on Windows double-click Docker Desktop Installer.exe · then open the Docker Desktop app and leave it running. No curl command needed

    Open the Docker download page ↗

    When it’s done, close Terminal, open it again and run the same command. You should see a version number: docker --version

    Install once; it works for everything that needs Docker

  2. Open Terminal

    Mac: press Command+Space and type Terminal · Windows: open PowerShell

  3. Paste this command and press Enter

    claude mcp add --scope user unified-ai-system -- docker run -i --rm ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0
  4. Type /mcp in Claude Code

    You should see Unified AI System MCP Gateway listed as connected

  5. Try it

    Use Unified AI System MCP Gateway and tell me what you can do for me, then show me one thing

    ✓ If Claude Code says it is using Unified AI System MCP Gateway or shows a Unified AI System MCP Gateway tool in the answer, you are connected

Menu names checked against Anthropic’s help page on Nov 11, 2026 · your screen may differ slightly by app version

Works with Cursor, with some setup first

Install Docker on your computer first (one time)

Let AI install it

Copy this message into a Cursor chat. The AI checks your computer, tells you which file it will change, and waits for your OK every time.

Please install the MCP server Unified AI System MCP Gateway for Cursor on this computer, in this order: 1) Check Docker with docker --version. If it is missing, tell me to install it from https://www.docker.com/products/docker-desktop/; do not install it yourself. 2) Use only this config (already checked by PopNx MCP): { "mcpServers": { "unified-ai-system": { "command": "docker", "args": [ "run", "-i", "--rm", "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0" ] } } } 3) Tell me the full path of the file you will change, or the command you will run, and wait for my OK. 4) No key is needed. 5) Do what we agreed, then check that the Unified AI System MCP Gateway tools really show up.

This message has no key in it. If a key is needed, the AI will ask you to paste it yourself into the file or the app’s field.

  1. Install Docker (if you don’t have it)

    Already installed? Check first · Open Terminal, type this command and press Enter

    docker --version
    • A version number such as Docker version … = installed, skip this step
    • "command not found" (on Windows: "is not recognized") = not installed yet, or installed but Terminal needs to be closed and reopened
    How to open Terminal

    Mac: press ⌘ + Space, type Terminal, press Enter

    Windows: open Start, type Windows PowerShell, then Open

    Not installed: Install Docker Desktop: on Mac open Docker.dmg and drag the icon to Applications (pick Apple silicon or Intel to match your Mac) · on Windows double-click Docker Desktop Installer.exe · then open the Docker Desktop app and leave it running. No curl command needed

    Open the Docker download page ↗

    When it’s done, close Terminal, open it again and run the same command. You should see a version number: docker --version

    Install once; it works for everything that needs Docker

  2. Click this button to open Cursor

    Open in Cursor

    Cursor asks you to confirm; click Install

  3. Open the AI chat in Cursor in Agent mode

  4. Try it

    Use Unified AI System MCP Gateway and tell me what you can do for me, then show me one thing

    ✓ If Cursor says it is using Unified AI System MCP Gateway or shows a Unified AI System MCP Gateway tool in the answer, you are connected

Menu names checked against Cursor’s help page on Nov 11, 2026 · your screen may differ slightly by app version

Works with VS Code, with some setup first

Install Docker on your computer first (one time)

Let AI install it

Copy this message into a VS Code chat. The AI checks your computer, tells you which file it will change, and waits for your OK every time.

Please install the MCP server Unified AI System MCP Gateway for VS Code on this computer, in this order: 1) Check Docker with docker --version. If it is missing, tell me to install it from https://www.docker.com/products/docker-desktop/; do not install it yourself. 2) Use only this config (already checked by PopNx MCP): { "servers": { "unified-ai-system": { "type": "stdio", "command": "docker", "args": [ "run", "-i", "--rm", "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0" ] } } } 3) Tell me the full path of the file you will change, or the command you will run, and wait for my OK. 4) No key is needed. 5) Do what we agreed, then check that the Unified AI System MCP Gateway tools really show up.

This message has no key in it. If a key is needed, the AI will ask you to paste it yourself into the file or the app’s field.

  1. Install Docker (if you don’t have it)

    Already installed? Check first · Open Terminal, type this command and press Enter

    docker --version
    • A version number such as Docker version … = installed, skip this step
    • "command not found" (on Windows: "is not recognized") = not installed yet, or installed but Terminal needs to be closed and reopened
    How to open Terminal

    Mac: press ⌘ + Space, type Terminal, press Enter

    Windows: open Start, type Windows PowerShell, then Open

    Not installed: Install Docker Desktop: on Mac open Docker.dmg and drag the icon to Applications (pick Apple silicon or Intel to match your Mac) · on Windows double-click Docker Desktop Installer.exe · then open the Docker Desktop app and leave it running. No curl command needed

    Open the Docker download page ↗

    When it’s done, close Terminal, open it again and run the same command. You should see a version number: docker --version

    Install once; it works for everything that needs Docker

  2. Click this button to open VS Code

    Open in VS Code

    VS Code asks you to confirm; click Install

  3. Open the AI chat in VS Code in Agent mode

  4. Try it

    Use Unified AI System MCP Gateway and tell me what you can do for me, then show me one thing

    ✓ If VS Code says it is using Unified AI System MCP Gateway or shows a Unified AI System MCP Gateway tool in the answer, you are connected

Menu names checked against VS Code’s help page on Nov 11, 2026 · your screen may differ slightly by app version

Other apps that use MCP, such as Windsurf, Zed or Gemini CLI, each add servers differently. See how in the app list.

See every app that uses MCP →

JSON settings are under “For developers” below

Stuck? Ask people who use Unified AI System MCP Gateway💬 Discussion →
For developers: Terminal commands and JSON settings

Installed on your computer

Claude Code

claude mcp add --scope user unified-ai-system -- docker run -i --rm ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0

Claude Desktop (claude_desktop_config.json)

{
  "mcpServers": {
    "unified-ai-system": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0"
      ]
    }
  }
}

Cursor (~/.cursor/mcp.json)

{
  "mcpServers": {
    "unified-ai-system": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0"
      ]
    }
  }
}

VS Code (mcp.json)

{
  "servers": {
    "unified-ai-system": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0"
      ]
    }
  }
}

These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.

At a glance

GitHub starsi
10 (Stars for the whole repo)
This repo holds several MCP listings; its stars count once in rankings
Downloads per weeki
Not on npm/PyPI
Licencei
Apache-2.0
Connection
Runs on your computer
Latest versioni
0.8.0
Published to registryi
Sep 25, 2026
Code last changedi
Nov 19, 2026
Maker
happy520ai
Packages
oci: ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0
Registry status
Active

Which apps can use it

Based on how this server connects and on each app’s own docs, checked Nov 10, 2026.

Run on your computer in

ClaudeGooseLM StudioMsty StudioCherry StudioCursorVS Code + GitHub CopilotVisual StudioWindsurf (Devin Desktop)ZedKiroClineContinueJetBrains AI AssistantClaude CodeCodex CLIGemini CLIOpenCodenodecodeWarpLibreChatRaycast

All apps, with pros and cons →

Related guides

What it can do

Runs on your computer; the tool list appears in your AI app after you install it.

Basic checksi

We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 19, 2026

  • ✓ Pass

    Maker matches the repo

    The registry name was verified with the GitHub account happy520ai, which owns the repo

  • ✓ Pass

    Still maintained

    Code last changed today

  • ✓ Pass

    Has a licence

    Uses Apache-2.0

  • – Not applicable

    Endpoint responds

    This one runs on your computer; there is no internet endpoint

  • – Not applicable

    Sign-in with OAuth

    Runs on your computer and uses local settings instead

  • … Not checked yet

    No hidden instructions in tool descriptions

    Runs on your computer; we can only read tool lists from internet endpoints

Published vulnerabilities (OSV.dev)

Not checked: only npm and PyPI packages with a version can be checked

What we check → · JSON

Makers: ask for a re-check or report wrong info
For makers: add a badge to your README

If you made this, paste this code into your README. The badge links back to this page.

[![Featured on PopNx MCP](https://mcp.popnx.com/badge/happy520ai-unified-ai-system.svg)](https://mcp.popnx.com/en/servers/happy520ai-unified-ai-system)

Discussion

Ask, share a tip or report that it does not work. Sorted by what people marked helpful.

No posts yet. Be the first to ask or share.

Posts here are written by users, not the PopNx team.

Sign in to post

User reviews

No reviews yet

Sign in to write a review

Compared with others in Search & AI memoryi

  • More stars than 83% of the other 6416 servers in this category
  • Published to the registry more recently than 41% of the other 6416 servers in this category
  • Basic checks equal to or better than 99% of the other 6416 servers in this category

Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.

Common questions

Is Unified AI System MCP Gateway safe?

It passed 3 of 3 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.

What keys or codes does it need?

The registry does not list a required key, though some servers ask you to sign in on first use.

Does it run on my computer or in the cloud?

It runs on your computer. Your AI app starts it when needed, and data does not pass through the maker’s servers (unless it calls an online service itself).

Is it still maintained?

The GitHub repo last changed today, and the latest version (0.8.0) reached the registry 2 months ago.

Which apps can use it?

Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. You also need Docker on your computer.

What can I use instead?

In the same category: OpenSEO, Exa, Microsoft Learn MCP, Docs, COTAL actions (cotal.ai)

Similar alternatives

Same category, other makers, commands already checked. Official ones first, then by stars.

Reading a card:↓ Downloads per weeki★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi

OpenSEO

Keyword research, site audits, rank tracking and AI visibility for your AI agent.

by every-appWidely used★ 28.2KUpdated 5 days ago

Runs in the cloud✓ everyapp.com

Exa

Connect AI agents to Exa for web search, content fetching, and multi-step research.

by exa-labsWidely used★ 5.2KUpdated 2 months ago

Runs in the cloud✓ service owner · exa.aiTried by the PopNx team

Microsoft Learn MCP

Official Microsoft Learn MCP Server – real-time, trusted docs & code samples for AI and LLMs.

by MicrosoftDocsWidely used★ 2KUpdated 12 months ago

Runs in the cloud✓ service owner · microsoft.comTried by the PopNx team

Docs

Search and read Empryo's documentation. Read-only, no auth, no local access.

by proxysoulGrowing★ 1.6KUpdated 2 months ago

Runs in the cloud✓ empryo.com
Compared with similar ones
This one
Unified AI System MCP Gateway
OpenSEOExaMicrosoft Learn MCPDocsCOTAL actions (cotal.ai)
Runs whereInstalls on your computerRuns in the cloud, nothing to installRuns in the cloud, nothing to installRuns in the cloud, nothing to installRuns in the cloud, nothing to installRuns in the cloud, nothing to install
Needs a keyNoNoNoNoNoNo
Add with + in ClaudeNoYesYesYesNoNo
Basic checks3/35/52/44/44/44/4
LicenseApache-2.0MITMITCC-BY-4.0NOASSERTIONApache-2.0
PriceNo cost listed in the registryNo cost listed in the registryNo cost listed in the registryNo cost listed in the registryNo cost listed in the registryNo cost listed in the registry

README from the maker

Open the README

Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.

Unified AI System MCP Server

Connect Codex and other MCP hosts to the credential-free Unified AI System preview over stdio or MCP Streamable HTTP.

The server starts an isolated local gateway automatically, pins it to the deterministic fake provider, exposes authenticated permission-scoped tools, and removes the gateway when the MCP session ends. It does not enable or authorize real provider calls.

Tools

ToolBehavior
gateway_healthInspect gateway health and provider safety state.
gateway_readinessInspect first-run and chat readiness.
agent_governance_statusInspect Agent Governance status when the authenticated identity has platform-status access.
agent_governance_listList only governed Agents visible to the authenticated tenant.
agent_governance_describeDescribe one Agent only when it belongs to the authenticated tenant.
gateway_prompt_enhanceStructure a natural-language request locally without a provider call.
gateway_prompt_enhance_llmEnhance prompts via optional LLM backend (or falls back when unavailable).
gateway_chatSend one fake-provider-only chat request.
knowledge_readinessInspect local knowledge infrastructure.
knowledge_retrieveSearch the local knowledge base by keyword.
workflow_healthInspect the governed workflow subsystem.
workflow_actionsList workflow action definitions.
workflow_runExecute a three-step local workflow and atomically publish a tenant-partitioned artifact. Managed mode injects a dedicated file_write Agent; external governed gateways require agentId.
workforce_healthInspect the workforce subsystem.
workforce_agentsList workforce agent descriptors.

Agent Governance inspection tools are read-only. workflow_run writes one controlled local artifact. Agent creation remains available through the REST, shared-SDK, and human Agent Console surfaces because the MCP tool protocol cannot yet prove idempotent, cancellation-safe generation after a transport timeout. No MCP tool exposes Agent creation/execution/revocation, approval/rejection, or policy creation/activation. Governance tools never accept tenantId or ownerUserId: the Gateway binds both from its authenticated identity, and cross-tenant descriptions fail as not found. workflow_run never overwrites an existing path and is routed through Agent Governance when the connected Gateway enables that control plane. The chat tool checks the gateway safety state before every request and fails closed unless realProviderEnabled is exactly false and the response proves executionMode: "fake".

The source build exposes sixteen tools, including the four Agent Governance lifecycle tools and the provider-free gateway_prompt_enhance preview. Published images expose the tool set shipped by their pinned tag.

Run From Source

From the repository root:

pnpm install --frozen-lockfile

node packages/mcp-server/src/index.js

The project-level .codex/config.toml registers that command automatically in trusted Codex projects. Restart Codex after cloning or changing MCP configuration, then use /mcp to inspect the server.

Streamable HTTP

Start the source-build HTTP transport:

pnpm mcp:http

The MCP endpoint is http://127.0.0.1:3210/mcp. It listens on loopback by default, validates Host and Origin, starts the same fake-provider gateway, and exposes the same sixteen source-build tools as stdio. Point any MCP Streamable HTTP client at that URL.

For a local token-protected endpoint, set MCP_HTTP_AUTH_TOKEN and send it as Authorization: Bearer . Binding beyond loopback fails closed unless all of these variables are present:

VariableRequirement
MCP_HTTP_HOSTNon-loopback bind address, such as 0.0.0.0.
MCP_HTTP_AUTH_TOKENAt least 32 bytes; never place it in source control.
MCP_HTTP_ALLOWED_HOSTSComma-separated request hostnames, without ports.
MCP_HTTP_ALLOWED_ORIGINSComma-separated browser origin hostnames, without schemes or ports.

Example for a private deployment behind a TLS reverse proxy:

MCP_HTTP_HOST=0.0.0.0 \

MCP_HTTP_AUTH_TOKEN=replace-with-a-random-secret-of-at-least-32-bytes \ MCP_HTTP_ALLOWED_HOSTS=mcp.example.com \ MCP_HTTP_ALLOWED_ORIGINS=console.example.com \ pnpm mcp:http

The built-in Bearer token is a bounded self-hosting control, not an OAuth authorization server. Terminate TLS at a trusted reverse proxy and use a dedicated identity layer before exposing the endpoint to untrusted networks. The published v0.5.0 container remains stdio-only; the HTTP command above is source-build functionality until a later release publishes it.

Add To Codex With Docker

No clone or API key is required:

codex mcp add unified-ai-system -- docker run --rm -i ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0

The dedicated image starts the MCP server by default; no command override is required, and it invokes Node directly so package-manager output cannot pollute the JSON-RPC stream. Its OCI identity and stdio transport are declared in the repository root server.json for the official MCP Registry.

The equivalent config.toml entry is:

[mcp_servers.unified_ai_system]

command = "docker" args = [ "run", "--rm", "-i", "ghcr.io/happy520ai/unified-ai-system/mcp-server:0.8.0", ] startup_timeout_sec = 45 tool_timeout_sec = 60 default_tools_approval_mode = "writes"

For clients that accept the standard JSON mcpServers shape, use the generic MCP client configuration guide.

Try It In Codex

Restart Codex after adding the server, then run /mcp verbose to inspect the connected tools. A useful first task is:

Use gateway_prompt_enhance to turn "build a small API for my team" into a

coding prompt. Show the preserved original request, detected language,

enhanced prompt, and proof that no provider was called.

Then prove the fake-chat safety boundary:

Use the Unified AI System MCP tools to check gateway health and readiness,

then send MCP_READY through gateway chat only if the gateway proves

fake-only mode. Report the provider, model, execution mode, and response.

The 60-second Codex MCP quickstart includes two more safe tasks, expected evidence, diagnostics, and the removal command.

Connect To An Existing Safe Gateway

Set AI_GATEWAY_MCP_URL and a dedicated gateway access token to use an already running instance:

AI_GATEWAY_MCP_URL=http://127.0.0.1:3100 \

AI_GATEWAY_MCP_AUTH_TOKEN= \ node packages/mcp-server/src/index.js

The managed gateway uses a private least-privilege token whose lifetime is bounded by the isolated child process; it is never emitted in MCP results. External gateway startup rejects missing or weak tokens, URL credentials, non-loopback plaintext HTTP, invalid authentication, or any gateway that may call a real provider. Provider credentials and real-provider execution remain outside this preview surface.

For Agent Governance, give the dedicated Gateway identity only dashboard:read and workflow:run unless a broader operator workflow is explicitly required. agent_governance_status additionally requires the Gateway's platform-tenant identity; ordinary tenants can still use tenant-scoped list, generate, and describe. Do not grant workflow:approve or user:admin to an MCP model merely to use these tools.

Verify

The tests launch both transports through the official MCP v2 client. They pin the modern 2026-07-28 stateless era on stdio and Streamable HTTP, retain legacy 2025-11-25 plus 2025-06-18 initialize compatibility, list all source-build tools, prove local prompt enhancement and fake-provider safety, exercise HTTP/CORS access controls, close the clients, and confirm that each managed gateway stopped. The container-safe raw smoke also uses server/discover plus a per-request _meta envelope rather than inferring modern support from the SDK version:

pnpm verify:mcp