Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Description from the maker (original English)
1 MCP server, 13 databases, zero config. Multi-protocol transport with Auth + RBAC.
These three are never merged into one score: opinions, measurements and popularity are different things.
We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.
Pick the AI app you use, then copy the settings.
You need
Which app do you use?
Pick your app above. We remember it, so next time the button says “Install in …”.
Open Claude Desktop → Settings → Developer → Edit Config, and paste this into claude_desktop_config.json
{
"mcpServers": {
"mosta-net": {
"command": "npx",
"args": [
"-y",
"@mostajs/net"
]
}
}
}Run this in your terminal
claude mcp add mosta-net -- npx -y @mostajs/netPaste into .cursor/mcp.json in your project, or ~/.cursor/mcp.json
{
"mcpServers": {
"mosta-net": {
"command": "npx",
"args": [
"-y",
"@mostajs/net"
]
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
Paste into .vscode/mcp.json in your project
{
"servers": {
"mosta-net": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@mostajs/net"
]
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
This one runs on your computer, and ChatGPT can only connect to servers on the internet. Pick another app.
Your choice stays on this device in a cookie; no account needed.
These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.
Runs on your computer; the tool list appears in your AI app after you install it.
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 7, 2026
The registry name was verified with the GitHub account apolocine, which owns the repo
We have not read this repo from GitHub yet
We have not read this repo from GitHub yet
This one runs on your computer; there is no internet endpoint
Runs on your computer and uses local settings instead
Runs on your computer; we can only read tool lists from internet endpoints
Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.
It passed 1 of 1 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.
The registry does not list a required key, though some servers ask you to sign in on first use.
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).
The GitHub repo last changed –, and the latest version (2.0.42) reached the registry 7 months ago.
Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. You also need Node.js on your computer.
In the same category: Spec Driven Development, Remote Desktop Commander, First Sale Scout
Servers in the same category from other makers, sorted by stars.
Reading a card:★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Playbooks and a 14-day checklist for founders winning their first paying customers.
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.
One schema, 11 transports, 13 databases, multi-tenant auth — out of the box.
Schema-driven multi-protocol API server for the @mostajs ecosystem. Built with TypeScript on Fastify, designed for polyglot consumption (14 native NetClients).
Author : Dr Hamid MADANI \ Homepage : octonet.amia.fr · mcp.amia.fr
The socket board.
@mostajs/ormconnects to 13 databases; Octonet
(
@mostajs/net) re-exposes the same entities over 11 transports (REST,
WebSocket, gRPC, MQTT, AMQP…); NetClient then fans out to **18 language
clients** — so any external app plugs into your data, in any runtime.
Octonet (@mostajs/net) is a Node.js multi-transport server that exposes a single @mostajs/orm schema as 11 different network protocols simultaneously : REST, GraphQL, WebSocket, SSE, JSON-RPC, MCP, gRPC, tRPC, OData, NATS, Arrow Flight.
The same User entity is reachable as :
curl https://octonet.amia.fr/api/v1/User # REST
curl https://octonet.amia.fr/graphql -d '{"query":...}' # GraphQL wscat -c wss://octonet.amia.fr/ws # WebSocket curl -N https://octonet.amia.fr/mcp # MCP (Claude Desktop)
Backed by @mostajs/orm (13 SGBD dialects). Backed by @mostajs/rbac + @mostajs/auth + @mostajs/api-keys for multi-tenant identity. Backed by @mostajs/mproject for multi-project routing.
Octonet is the second cerebral lobe of the @mostajs trilogy :
The fastest way to test Octonet from any of the 14 runtimes :
alice-42)curl -X POST https://octonet.amia.fr/try \
-H "Content-Type: application/json" \ -d '{"alias":"alice-42"}'
{ "status": "ok", "data": { "alias": "alice-42", "projectSlug": "sandbox-alice-42", "apiKey": "sk_test_…(64 chars, shown ONCE)", "permissions": { "projects":["sandbox-alice-42"], "operations":["read","write"], "transports":["rest","mcp"] }, "expiresAt": "2026-05-02T18:38:19.920Z", "quota": { "reqPerDay": 500 }, "exampleCurl": "curl https://octonet.amia.fr/api/v1/sandbox-alice-42/User -H \"X-API-Key: sk_test_…\"", "mcpUrl": "https://octonet.amia.fr/mcp" } }
Your sandbox is :
User, Product, Order entitiesUse the apikey in any of the 14 NetClients :
# Java / .NET / Python / Go / etc.
export MOSTAJS_NET_URL=https://octonet.amia.fr export MOSTAJS_NET_API_KEY=sk_test_…
| Tier | Cible | Auth | Quota | Use case |
|---|---|---|---|---|
T1 — Sandbox publique /try | "I want to test in 10 minutes" | alias seul (pas d'email) | 500 req/jour, TTL 7j | démo NetClients, intégration tests |
| T2 — Compte enregistré octocloud.amia.fr | "j'utilise vraiment, gratuit" | email + mot de passe | 10 000 req/jour | apps personnelles, side-projects |
T3 — Self-host npx @mostajs/net init | "tes données chez toi" | admin local | aucune | enterprise, on-premise, AGPL |
Auto-generated from your registered schemas. Toggle each via env var (default : all enabled).
| # | Transport | Endpoint | Use case | Notes |
|---|---|---|---|---|
| 1 | REST | /api/v1/{Entity} · /api/v1/{project}/{Entity} | universel | 15 routes par entité (CRUD + count + search + aggregate + bulk + relations) |
| 2 | GraphQL | /graphql (POST) | front-end riches | schéma + GraphiQL IDE auto-générés via mercurius |
| 3 | WebSocket | wss://…/ws | temps réel | events entity.created/updated/deleted/upserted broadcast |
| 4 | SSE | /events (GET stream) | mobile / browser-friendly | server-sent events |
| 5 | JSON-RPC | /rpc (POST) | EVM-adjacent, classic | JSON-RPC 2.0 + method discovery |
| 6 | MCP | /mcp (POST/GET SSE) | agents IA (Claude, ChatGPT) | 15 tools/entité auto-générés (listed on mcp.so) |
| 7 | gRPC | :50051 | inter-services low-latency | .proto auto-généré, 6 RPCs/entité |
| 8 | tRPC | /trpc/{Entity}.{op} | TypeScript fullstack | type generation côté client |
| 9 | OData | /odata/{Collection} (+ $metadata) | SAP/Microsoft Dynamics | OData v4 ($filter, $select, $orderby) |
| 10 | NATS | mostajs.{Entity}.{op} | pub/sub edge | request-reply messaging |
| 11 | Arrow Flight | /arrow/* | analytics columnaire | streaming zero-copy |
Le même schéma, 11 portes d'entrée différentes, 0 codegen.
Persistence dialects fournis par @mostajs/orm :
| Catégorie | Bases | Dialect ID |
|---|---|---|
| SQL mainstream | PostgreSQL, MySQL, MariaDB, SQLite | postgres, mysql, mariadb, sqlite |
| SQL enterprise | Oracle, SQL Server, DB2, SAP HANA, HSQLDB, Sybase | oracle, mssql, db2, hana, hsqldb, sybase |
| NewSQL / Cloud | CockroachDB, Google Cloud Spanner | cockroachdb, spanner |
| NoSQL | MongoDB | mongodb |
Switch dialect = changer 1 ligne d'env :
DB_DIALECT=postgres → DB_DIALECT=mongodb
SGBD_URI=postgresql://… → SGBD_URI=mongodb://…
Octonet utilise un middleware d'authentification basé sur API keys avec scopes (orienté machine-to-machine) + RBAC pour la gestion humaine. L'orchestrateur agnostique vit dans @mostajs/auth/lib/check-request.ts.
HTTP Request
│ ▼ authGuard (Fastify adapter) │ ▼ checkRequest (@mostajs/auth — framework-agnostic) │ ├─ extract X-API-Key from header / Bearer token / ?apikey= ├─ resolveApiKey (@mostajs/api-keys) — DB lookup + bcrypt verify ├─ isScopeAuthorized (@mostajs/api-keys) — check scope.values │ for each (scope, value) pair : │ checks: [{scope:'projects', value:slug}, │ {scope:'operations', value:'read'|'write'|'admin'}, │ {scope:'transports', value:'rest'|'mcp'|...}] ├─ touchApiKey — fire-and-forget : lastUsedAt, usageCount, lastIp │ ▼ ormHandler (entité CRUD) ─→ sanitizer (strip password/hash/tokens) │ ▼ HTTP Response
Une apikey a la shape :
{
permissions: { scopes: { projects: ['my-project','demo'] | '*', // declared by @mostajs/mproject operations: ['read','write','admin'] | '*', // declared by @mostajs/orm transports: ['rest','graphql','mcp'] | '*', // declared by @mostajs/net // anything: any module can register a new scope }, rateLimit: 500, } }
@mostajs/api-keys ne connaît PAS les noms projects/operations/transports — c'est volontaire. Chaque module enregistre ses scopes au boot via registerScope(dialect, {name, label, …}). Le check est générique : isScopeAuthorized(perms, scope, value).
| Cas | HTTP code | Body |
|---|---|---|
| Pas d'apikey | 401 | {error: {code:'UNAUTHORIZED', message:'API key required…'}} |
| Apikey invalide / révoquée | 401 | {error: {code:'UNAUTHORIZED', message:'Invalid or revoked API key'}} |
| Apikey valide hors scope | 403 | {error: {code:'FORBIDDEN', message:'API key not authorized for X="Y"'}} |
| Apikey OK | 200/201 | {status:'ok', data: …} (sanitized — no password/hash) |
L'endpoint /mcp autorise un fallback automatique sur l'apikey labelée public-default quand aucune clé n'est présentée — préserve la compat avec mcp.so / Claude Desktop. La clé publique est read-only sur le projet `default` uniquement → écritures bloquées.
Les champs password, hash, verifyToken, resetToken, apiKeyHash, secret, privateKey sont automatiquement strippés de toutes les réponses JSON par un middleware global. Aucun risque de leak à travers /api/v1/User?limit=1.
Le catalogue des scopes vit en base de données du projet accueillant (pas de fichier JSON, pas de constantes hardcodées) :
-- Auto-créées au boot via @mostajs/api-keys
CREATE TABLE api_key_scopes ( id, name, label, description, icon, cardinality, valuesSource ('static' | 'dynamic'), dynamicSourceRef );
CREATE TABLE api_key_scope_values ( id, scopeName, value, label, sortOrder, metadata );
Chaque module enregistre ses scopes au boot :
import { registerScope } from '@mostajs/api-keys/server'
// Dans mosta-net/src/server.ts (boot) await registerScope(dialect, { name: 'transports', label: 'Network transports', cardinality: 'low', valuesSource: 'static', staticValues: [ {value:'rest', sortOrder:1}, {value:'graphql', sortOrder:2}, // … 9 more ], })
await registerScope(dialect, { name: 'projects', label: 'Projects', cardinality: 'high', valuesSource: 'dynamic', dynamicSourceRef: 'Project.slug', // queried at runtime })
L'admin UI charge le catalogue dynamique via GET /api/api-keys/scopes et rend une matrice (composant React ApiKeyScopeMatrix dans @mostajs/api-keys/components/).
Via @mostajs/mproject — N bases de données isolées sur le même serveur.
# Default project (DB_DIALECT + SGBD_URI au boot)
curl http://localhost:4488/api/v1/User
curl http://localhost:4488/api/v1/analytics/events
curl http://localhost:4488/api/v1/events -H "X-Project: analytics"
curl -X POST http://localhost:4488/api/projects \
-H "X-API-Key: " \ -d '{ "name": "analytics", "dialect": "mongodb", "uri": "mongodb://localhost:27017/analytics",
Only the beginning is shown; read the rest in the repo