Agentgates Compute

ai.agentgatesPublished by agentgates.aiOfficial publisherv1.0.3Updated Nov 5, 2026

Confidential computers for AI swarms: sealed VMs (Intel TDX) by the hour, USDC over x402.

The publisher's own server: Agentgates Wallet

VerifiedStreamable HTTPWeb executableCloud & InfrastructureFinance

Overview

AI-generated overview

Read-only MCP server for Agentgates confidential VMs: lists hourly plans, quotes exact USDC prices over x402, and serves setup docs.

What you can do
  • List confidential VM plans with vCPU, memory, disk, traffic and live USDC prices per payment network
  • Get an exact 402 quote for a plan, hours or minutes, or a renewal, including what to sign
  • Read the live compute docs on paying, polling a machine until it runs, renewing and signing owner calls
  • These tools are free and read only; paying happens outside the server with your own wallet
When to use it
Use it when an assistant needs to price or plan renting a confidential Intel TDX virtual machine paid in USDC over x402, or needs the current payment and lifecycle instructions before a purchase. Not needed if you only want to browse general Agentgates products.
Requirements
Remote Streamable HTTP endpoint at mcp.agentgates.ai; no account, sign-up or API key for these read-only tools. Network access to the Agentgates API. Actually renting a machine requires your own wallet with USDC and an x402 payment header.
Before you install
The server itself is read only, but the documented workflow leads to real spending: machines are paid in USDC by the paying wallet, and the README warns everything costs its owner money. Set a budget and confirm before signing any payment; never sign a second payment if an answer is unclear. Wallet tools with money-moving proposals live on a separate OAuth server.

Installation

In SourceWeft

  1. Open Agentgates Compute in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/ai.agentgates%2Fcompute/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "agentgates-compute": {
      "type": "http",
      "url": "https://mcp.agentgates.ai/compute"
    }
  }
}

README

[Agentgates]

Agentgates

Confidential computers for AI swarms, models, phone calls, inboxes and an x402 directory, sold to AI agents for USDC over x402. No signup and no API key: the wallet that pays is the account.

agentgates.ai | Prices, read live | llms.txt

What agents can buy

ProductWhat it isContract
ComputeConfidential computers for AI swarms: a dedicated confidential VM (Intel TDX) for each agent, for trading swarms or any other swarm, by the hour, running your Docker image or compose file, with a remote-attestation quotedocs, catalog
InferenceModels in the standard chat format, paid by the request: confidential models in attested hardware, and a cheaper open tier without attestationdocs, catalog
PhoneThe payphone: one real call with no account, in your own voice over a websocket or with our operator. Or a number of your owndocs
VoiceOur AI operator makes the call and answers with the transcript and the outcomedocs
EmailInboxes for agents: receive mail and sign-up codes, and senddocs
x402 directoryFind, review and list x402 servicesdocs

Install

Skills, for any coding agent:

bash
npx skills add agentgates-ai/agentgates

Claude Code:

/plugin marketplace add agentgates-ai/agentgates/plugin install agentgates@agentgates

Gemini CLI:

bash
gemini extensions install https://github.com/agentgates-ai/agentgates

MCP servers (Streamable HTTP), both in the official MCP registry:

ServerURLSign-inRegistry name
Computehttps://mcp.agentgates.ai/computeNone: plans, quotes and docsai.agentgates/compute
Wallethttps://mcp.agentgates.ai/api/mcpOAuth. Every money move is a proposal its owner approvesai.agentgates/mcp

CLI:

bash
npm i -g @agentgates/cliagentgates init

How paying works

  1. Send the request with no payment header. The answer is a 402 naming the exact USDC amount on each network it accepts. Asking is free: nothing runs and nothing is charged.
  2. Sign that amount with the paying wallet.
  3. Send the same request again with the payment in X-PAYMENT (x402 v1) or PAYMENT-SIGNATURE (x402 v2, on compute today).

Or prepay a balance once and spend it request by request, signed by the wallet or with a bearer key.

Prices move with gas and with the providers, so they live in the 402 and the catalogs, never in this repo.

Examples

Machine-readable

WhatURL
Index for agents/llms.txt
Every contract in one file/llms-full.txt
OpenAPI/openapi.json, /openapi.yaml
A2A agent card and JSON-RPC door/.well-known/agent-card.json, POST https://agentgates.ai/a2a
Agent skills index/.well-known/agent-skills/index.json
ARD catalog/.well-known/ard.json
API catalog (RFC 9727)/.well-known/api-catalog
MCP server card/.well-known/mcp.json
How agents authenticate/auth.md
Prices/pricing

Spending

Everything here costs its owner money. An agent should pay only what its owner allowed: ask first, or stay inside a budget the owner set. Every skill in skills/ says so.

This repo

skills/ holds the same SKILL.md files agentgates.ai serves. .claude-plugin/ and .mcp.json make it a Claude Code plugin, gemini-extension.json a Gemini CLI extension, and registry/ holds the MCP registry records. The services themselves run at agentgates.ai.

Terms, Privacy, security.txt. Support: [email protected].

License

MIT

Source: README.md at commit d568cb4

Tools

3

compute_plans

Confidential VM plans and prices

Lists the confidential virtual machines (Intel TDX, memory encrypted in use) Agentgates rents to agents by the hour: each plan's vCPU, memory, disk and traffic, the hours one payment may buy, and the live price of each block of hours on each payment network (USDC over x402). Read from the live catalog at call time. Free: no account, no wallet, no payment. Pass `plan` to read one plan.

Parametersplan

compute_quote

Price one machine (the x402 offer)

Gets the exact price of one machine without paying: sends the create request with no payment and returns the API's HTTP 402 offer, one leg per network, each with the amount in USDC, the address it is paid to, the token, how long a signature stays good, and what to sign. The API checks the plan, the hours and the setup first, so a refusal costs nothing. Pass `machine_id` and `hours` instead to price a renewal of a running machine (any wallet may pay one). Where the API sells quarter-hours (compute_plans shows minutePrices), send `minutes` (15, 30 or 45) in place of `hours`. For a plan that lists regions (compute_plans shows them), send `region`. To buy, sign one leg with your own wallet and send the same request with an X-PAYMENT header (compute_docs has the whole contract).

Parametersplanporthoursimageregioncomposeminutesmachine_id

compute_docs

How to buy, poll, renew and control a machine

Returns the Agentgates compute docs, read live (markdown written for an agent to follow in one pass): the plans with live prices, how to pay the 402 on each network, how to poll a machine until it runs, how to renew it (any wallet may pay a renewal), how to sign owner calls (X-Wallet-Auth: EIP-191 over METHOD\npath\nnonce\ntimestamp, by the wallet that paid), the controls (stop, start, restart, resize, update), and what a lost answer or a refusal means.

Takes no parameters.

Version history

2
  1. v1.0.3LatestNov 5, 2026
  2. v1.0.2Nov 3, 2026

Score

100% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

Other Agentgates integrations