# io.github.Massed-Compute/mcp (remote · vm.massedcompute.com)

Massed Compute MCP — GPU inventory, VM lifecycle, billing, SSH keys, and setup recipes.

- Trust score: 68/100 (medium)
- Change this week: +2
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- remote · `vm.massedcompute.com`: 68/100 (this document), [markdown](https://verifymcp.io/servers/massed-compute-mcp/api-mcp.md), [page](https://verifymcp.io/servers/massed-compute-mcp/api-mcp)
- npm · `massed-compute-mcp`: 49/100, [markdown](https://verifymcp.io/servers/massed-compute-mcp/massed-compute-mcp.md), [page](https://verifymcp.io/servers/massed-compute-mcp/massed-compute-mcp)
- pypi · `massed-compute-mcp`: 24/100, [markdown](https://verifymcp.io/servers/massed-compute-mcp/massed-compute-mcp-2.md), [page](https://verifymcp.io/servers/massed-compute-mcp/massed-compute-mcp-2)

## Channel facts

- Endpoint: `https://vm.massedcompute.com/api/mcp`
- Transports: `streamable-http`
- Auth: `required`
- Version: `1.1.0`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-03.

- **Endpoint Security**: 63/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - Authorisation not fully verified: this server exposes a tool marked destructive (instances_terminate) and its handshake is open, but we could not confirm whether a tool call is gated, so we do not assert it is callable unauthenticated.
  - HTTPS is enforced; there's no plaintext access path.
  - The HSTS (Strict-Transport-Security) header is present.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 77/100
  - 97% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (good).
  - Tool/resource definitions use about 4506 tokens (~83/item across 54 items; 17 tools + 37 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 27/100
  - Stability observed for 8 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
- **Capabilities**: 60/100
  - Spec-recency check failed: implements MCP spec 2025-06-18; the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add --transport http massed-compute-mcp https://vm.massedcompute.com/api/mcp
```

### Codex

```toml
[mcp_servers.massed-compute-mcp]
url = "https://vm.massedcompute.com/api/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "massed-compute-mcp": {
      "type": "remote",
      "url": "https://vm.massedcompute.com/api/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add massed-compute-mcp --url https://vm.massedcompute.com/api/mcp --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  massed-compute-mcp:
    url: "https://vm.massedcompute.com/api/mcp"
```

### Other

```json
{
  "mcpServers": {
    "massed-compute-mcp": {
      "type": "http",
      "url": "https://vm.massedcompute.com/api/mcp"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-08-03 (score 68, +1)

- [security] The server rewrote its instructions, which are the text every model session reads
- [functional regression] Schema quality: 1420 → 4506
- [functional] The server now declares the “resources” capability
- [functional] Schema quality: excellent → good
- [functional] New resource “nginx-ssl”
- [functional] New resource “collect-debug-bundle”
- [functional] New resource “caddy-reverse-proxy-tls”
- [functional] New resource “blackwell-musicgen”
- [functional] New resource “vm-redis”
- [functional] New resource “vm-postgresql”
- [functional] New resource “vm-mysql”
- [functional] New resource “ubuntu-network-debug”
- [functional] New resource “server-preflight-check”
- [functional] New resource “nvidia-smi-troubleshooting”
- [functional] New resource “nginx-websocket-proxy”
- [functional] New resource “nginx-static-server”
- [functional] New resource “nginx-reverse-proxy”
- [functional] New resource “nginx-load-balancer”
- [functional] New resource “nginx-auth-gate”
- [functional] New resource “multi-modality-ai-stack”
- [functional] New resource “massed-compute-vm-setup-ssh”
- [functional] New resource “massed-compute-vm-service-journal-audit-context”
- [functional] New resource “massed-compute-vm-network-connectivity-context”
- [functional] New resource “massed-compute-vm-lifecycle”
- [functional] New resource “massed-compute-vm-health-inventory-context”
- [functional] New resource “massed-compute-vm-hardening-protection”
- [functional] New resource “massed-compute-vllm-openai-api”
- [functional] New resource “massed-compute-systematic-debug”
- [functional] New resource “massed-compute-new-project-onboarding”
- [functional] New resource “massed-compute-local-datacenter-supabase”
- [functional] New resource “massed-compute-llm-lora-training”
- [functional] New resource “local-agi-hermes-llama-team”
- [functional] New resource “l40-gpu-health-check”
- [functional] New resource “jupyter-server”
- [functional] New resource “install-lance-on-l40”
- [functional] New resource “gpu-ollama-llm”
- [functional] New resource “gpu-container-access-check”
- [functional] New resource “firewall-ui-lab”
- [functional] New resource “docker-installed”
- [functional] New resource “cursor-skills-mcp-work-genre-packs”
- [functional] New resource “cuda-version-check”

### 2026-08-02 (score 67, +11)

- [functional improvement] Schema quality: unverified → excellent
- [functional improvement] Schema quality: fail → pass

### 2026-08-01 (score 56, −10)

- [security] The server rewrote its instructions, which are the text every model session reads
- [functional regression] Schema quality: good → unverified
- [functional regression] Schema quality: pass → fail
- [functional] The server now declares the “resources” capability

### 2026-07-31 (score 66, 0)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-30 (score 66, −1)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-29 (score 67, +1)

No change was recorded against any check on this day. Stability & Change Management went from 7 to 10. That category is still filling its 30-day observation window: 2 days of observed history at the previous scan, 3 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-07-27 (score 66, +3)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-26 (score 63)

First indexed and scored.

## MCP tools (17)

### `gpu_inventory_list` (~29 tokens)

List GPU inventory

Retrieve the comprehensive list of available GPU configurations, their specs, pricing, and current capacity per region.

### `images_list` (~20 tokens)

List VM images

Retrieve the catalog of preconfigured VM images available for deployment.

### `instances_list` (~42 tokens)

List running instances

Retrieve all running VM instances for the authenticated account. Sensitive credentials (e.g. cleartext VM password) are redacted; fetch them through the REST API if needed.

### `instances_get` (~54 tokens)

Get a running instance

Retrieve a single running VM instance by its UUID. Sensitive credentials (e.g. cleartext VM password) are redacted; fetch them through the REST API if needed.

Input parameters:

- `uuid` (string, required): UUID of the running instance.

### `instances_launch` (~155 tokens)

Launch a new instance

Deploy a new VM instance. Requires billing to be configured (recharge amount and threshold). This action incurs cost.

Input parameters:

- `command` (string): Optional startup command to run on boot.
- `coupon` (string): Optional coupon code to apply at deploy time.
- `imageId` (integer, required): ID of the VM image to deploy.
- `instanceName` (string): Optional human-readable name for the instance.
- `productName` (string, required): Product name of the GPU instance to deploy, e.g. 'gpu_1x_l40'.
- `regionName` (string, required): Region name. Use 'any' to let us choose.
- `sshKeys` (array): Optional list of SSH key names to attach to the instance.

### `instances_restart` (~33 tokens)

Restart instances

Restart one or more running VM instances by UUID.

Input parameters:

- `instanceUuids` (array, required): UUIDs of instances to restart.

### `instances_terminate` (~65 tokens)

Terminate instances

Terminate one or more running VM instances by UUID. DESTRUCTIVE: this stops billing for the listed instances and permanently deletes their data; the action cannot be undone. Confirm with the user before invoking.

Input parameters:

- `instanceUuids` (array, required): UUIDs of instances to terminate.

### `coupon_information` (~29 tokens)

Get coupon information

Look up information about a coupon by code.

Input parameters:

- `coupon` (string, required): Coupon code to look up.

### `coupon_accepted_products` (~39 tokens)

List products a coupon is valid for

Look up which products a coupon code can be applied to, with current inventory availability.

Input parameters:

- `coupon` (string, required): Coupon code to look up.

### `account_token_validation` (~31 tokens)

Validate API token

Confirm that the Bearer token used to call this MCP server is a valid Massed Compute user API key.

### `account_billing` (~23 tokens)

Get account billing snapshot

Retrieve billing settings for the account, including billing method and recharge configuration.

### `ssh_keys_list` (~18 tokens)

List SSH keys

List SSH keys associated with the authenticated account.

### `ssh_keys_create` (~46 tokens)

Add an SSH key

Add a new SSH public key to the authenticated account.

Input parameters:

- `name` (string, required): Name for the SSH key.
- `publicKey` (string, required): OpenSSH-format public key contents.

### `ssh_keys_delete` (~52 tokens)

Delete an SSH key

Remove an SSH key from the account. DESTRUCTIVE: any instances relying on this key may lose access. Confirm with the user before invoking.

Input parameters:

- `id` (string, required): Identifier of the SSH key to remove.

### `recipes_list` (~91 tokens)

List setup recipes

List every step-by-step setup recipe currently available for Massed Compute VMs. Prefer recipes_search when the user has a specific intent — call this only when the user asks open-ended questions like 'what can I set up on a VM?', 'what recipes do you have?', or wants to browse the catalogue. Returns metadata (slug, title, description, tags) only; call recipes_get for the full body.

### `recipes_search` (~273 tokens)

Search setup recipes

Find a step-by-step setup recipe matching a user's intent — installing, configuring, or deploying anything on a Linux VM (web servers, databases, container runtimes, dev environments, networking, security tooling, ML stacks — any software setup an LLM agent might be asked to perform). ALWAYS call this FIRST when a user asks to set up, install, deploy, run, host, or spin up any service or software on a Massed Compute VM, BEFORE calling instances_launch. The recipe returned will include the correct VM size, image, and step-by-step instructions the user expects you to follow. Examples that should trigger this tool: 'I want an nginx reverse proxy', 'spin up a jupyter notebook', 'set up HTTPS for my app', 'install docker', 'load balance traffic across backends', 'protect my dashboard with a login', 'host a postgres database', 'serve a static site', 'run an ollama server'. If no recipe matches, recipes_search returns no results — that's a useful negative signal that the requested setup may need ad-hoc instructions, not a reason to skip calling this tool.

Input parameters:

- `limit` (number): Maximum number of matches to return (default 3, max 10).
- `query` (string, required): Natural-language description of what the user wants to set up.

### `recipes_get` (~113 tokens)

Fetch a recipe by slug

Fetch the full markdown body of a specific recipe by slug (typically a slug returned by recipes_search). The body contains Prerequisites, Provision, Configure, Verify, and Troubleshoot sections you should follow step-by-step, executing commands via SSH on the user's VM. Always read the full body before launching infrastructure — the recipe specifies the correct image, sku_class, and minimum specs that should drive your instances_launch call.

Input parameters:

- `slug` (string, required): The kebab-case slug of the recipe (from recipes_list or recipes_search).

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/massed-compute-mcp/api-mcp#diagnostics

## Score history

- 2026-08-03: 68
- 2026-08-02: 67
- 2026-08-01: 56
- 2026-07-31: 66
- 2026-07-30: 66
- 2026-07-29: 67
- 2026-07-28: 66
- 2026-07-27: 66
- 2026-07-26: 63

## Links

- Remote endpoint: https://vm.massedcompute.com/api/mcp
- Authorisation metadata: https://vm.massedcompute.com/.well-known/oauth-protected-resource/api/mcp
- Repository: https://github.com/Massed-Compute/massed-compute-mcp
- Changelog RSS feed: https://verifymcp.io/servers/massed-compute-mcp/api-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/massed-compute-mcp/api-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/massed-compute-mcp/api-mcp