How do I use Agent Skills API?

1. Click on "Deploy Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Agent Skills API search for react patterns rules" 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.

Agent Skills API

by agency-skills

Overview Schema Source Related Servers Discussions Admin

TypeScript

Local

Agency Skills — Backend API

Plataforma de gestión de agencias de agentes con sistema multi-agente, recursos personalizados (skills, rules, agentes, workflows) y protocolo MCP (Model Context Protocol) para integración con herramientas CLI.

Stack

Runtime: Node.js + TypeScript
Framework: NestJS 11 (Arquitectura Hexagonal + CQRS)
Base de datos: PostgreSQL (TypeORM 0.3)
Cache/Sesiones: Redis (ioredis)
Auth: Google OAuth 2.0 + JWT (access + refresh tokens)
API Docs: Swagger UI (/api)
MCP: @modelcontextprotocol/sdk (stdio + SSE)

Related MCP server: Spec Guard

Quick Start

# Instalar dependencias
pnpm install

# Configurar variables de entorno
cp .env.example .env
# Editar .env con tus credenciales (DB, JWT secrets, Google OAuth)

# Docker (PostgreSQL + Redis)
pnpm run docker:up

# Ejecutar migraciones
pnpm run db:migrate

# Desarrollo
pnpm run start:dev

El servidor arranca en http://localhost:8004.

Estructura

src/
├── core/                           # Domain + Ports (Clean Architecture)
│   ├── agents/                     # 17 agentes especializados
│   ├── domain/                     # Interfaces, entidades, tokens
│   ├── infrastructure/             # DB, adapters, MCP controller
│   ├── presentation/               # Controllers REST (health, rules, mcp, agents, sessions)
│   └── skills/                     # Sistema de skills Hermes
├── modules/
│   ├── auth/                       # Google OAuth + JWT (access/refresh)
│   ├── agencies/                   # CRUD de agencias
│   ├── agency-resources/           # Skills, rules, agents, workflows por agencia
│   ├── users/                      # Usuarios (lookup por IP)
│   ├── sessions/                   # Sesiones MCP
│   ├── issues/                     # Gestión de issues
│   ├── projects/                   # Detección de proyectos
│   ├── contexts/                   # Contextos de proyecto
│   ├── plans/                      # Planes de trabajo
│   ├── memory/                     # Sistema de memoria
│   └── agency-agents/              # Catálogo de agentes
├── agents/                         # Implementaciones de agentes
├── rules/                          # Reglas de código (archivos .rule.ts)
├── shared/                         # Middleware, utils, constantes
├── mcp-server.ts                   # Servidor MCP stdio
├── mcp-http-server.ts              # Servidor MCP HTTP
└── main.ts                         # Bootstrap NestJS

Variables de entorno

# Server
PORT=8004
CORS_ORIGIN=http://localhost:4200

# Database
DB_URL=postgres://user:password@localhost:5432/agent_skills?sslmode=disable
DB_SYNCHRONIZE=false

# Redis
REDIS_URL=redis://localhost:6379/0

# JWT
JWT_ACCESS_SECRET=your-secret-here
JWT_REFRESH_SECRET=your-other-secret-here
JWT_ACCESS_EXPIRES_IN=15m
JWT_REFRESH_EXPIRES_IN=7d

# Google OAuth
GOOGLE_CLIENT_ID=
GOOGLE_CLIENT_SECRET=

Ver .env.example para todas las variables disponibles (Exa AI, Context7, ChromaDB, etc.).

Scripts

# Desarrollo
pnpm run start:dev              # Hot-reload
pnpm run start:debug            # Debug mode

# Producción
pnpm run build                  # Build completo
pnpm run start:prod             # Ejecutar dist/main

# Base de datos
pnpm run db:migrate             # Ejecutar migraciones pendientes
pnpm run db:migrate:revert      # Revertir última migración
pnpm run db:generate            # Generar migración desde cambios en entidades

# MCP
pnpm run start:mcp              # MCP stdio server
pnpm run start:mcp-http         # MCP HTTP server
pnpm run build:mcp              # Build MCP stdio
pnpm run build:mcp-http         # Build MCP HTTP

# Tests
pnpm run test                   # Unit tests
pnpm run test:e2e               # E2E tests
pnpm run test:cov               # Coverage

# Docker
pnpm run docker:up              # Levantar servicios
pnpm run docker:down            # Parar servicios
pnpm run docker:logs            # Ver logs

# Lint
pnpm run lint                   # ESLint + auto-fix
pnpm run typecheck              # Type-check sin emitir

# Agentes CLI
pnpm run agent:setup-gemini     # Configurar Gemini MCP
pnpm run mcp:update-claude      # Actualizar config Claude Code

Endpoints principales

Método	Ruta	Descripción
`GET`	`/health`	Health check
`GET`	`/api`	Swagger UI
`GET`	`/mcp/sse`	MCP SSE endpoint
`POST`	`/mcp/message`	MCP message endpoint

Auth

Método	Ruta	Descripción
`GET`	`/api/auth/login/google`	Login con Google
`GET`	`/api/auth/refresh`	Renovar access token
`POST`	`/api/auth/logout`	Cerrar sesión

Agencies

Método	Ruta	Descripción
`POST`	`/v1/agency`	Crear agencia
`GET`	`/v1/agency/by-user/:userId`	Obtener agencia del usuario
`GET`	`/v1/agency/:id`	Obtener agencia por ID
`PATCH`	`/v1/agency/:id`	Actualizar agencia
`DELETE`	`/v1/agency/:id`	Eliminar agencia

Agency Resources

Método	Ruta	Descripción
`POST`	`/v1/agency/:agencyId/skills`	Crear skill
`GET`	`/v1/agency/:agencyId/skills`	Listar skills
`GET`	`/v1/agency/:agencyId/skills/:id`	Obtener skill
`PUT`	`/v1/agency/:agencyId/skills/:id`	Actualizar skill
`DELETE`	`/v1/agency/:agencyId/skills/:id`	Eliminar skill
		(ídem para `/rules`, `/agents`, `/workflows`)

Agency Members

Método	Ruta	Descripción
`GET`	`/v1/agency/:id/members`	Listar miembros
`POST`	`/v1/agency/:id/members`	Agregar miembro
`DELETE`	`/v1/agency/:id/members/:userId`	Remover miembro

Users

Método	Ruta	Descripción
`GET`	`/v1/users/search?q=`	Buscar usuarios por email/nombre
`GET`	`/v1/users/:id`	Obtener usuario por ID

Agentes (17)

Agente	Responsabilidad
RouterAgent	Orquestador principal, detecta intención y delega
SearchAgent	Búsqueda BM25 de reglas
IdentityAgent	Identidad MCP y prefijos
RulesAgent	Listado y gestión de reglas
CodeAgent	Generación de código
ArchitectureAgent	Validación arquitectónica
AnalysisAgent	Análisis de código
MetricsAgent	Métricas y tracking
PMAgent	Gestión de proyecto
IssueWorkflowAgent	Workflow de issues
GitHubAgent	Integración GitHub
FrontendArchitectureAgent	Arquitectura frontend
WebSearchAgent	Búsqueda web (Exa AI)
Context7Agent	Documentación de librerías
ContextAgent	Contexto de proyecto
ProjectHistoryAgent	Historial de proyecto
GraphifyAgent	Visualización de código
ObsidianAgent	Integración Obsidian

MCP (Model Context Protocol)

El servidor MCP expone herramientas a través de dos transportes:

stdio (pnpm run start:mcp): para integración con CLI locales (Claude Code, Gemini, OpenCode)
SSE (/mcp/sse + /mcp/message): para integración web y remota

Herramientas MCP

Tool	Descripción
`session_init_auto`	Inicializa sesión MCP con contexto del proyecto
`list_agency_skills`	Lista skills de la agencia activa
`get_skill_detail`	Detalle de un skill específico
`invoke_skill`	Ejecuta un skill

Arquitectura

Hexagonal (Ports & Adapters): lógica de negocio aislada de infraestructura
CQRS: comandos y queries separados con @nestjs/cqrs
Repository Pattern: interfaces en domain, implementaciones en infrastructure
JWT dual: access tokens (15min) en localStorage, refresh tokens (7d) en httpOnly cookie
IP-based auth: para conexiones SSE/MCP donde no hay cookie disponible

Documentación adicional

doc/ARCHITECTURE.md — Arquitectura completa del sistema
doc/MCP-CONFIG.md — Configuración MCP por herramienta
doc/IMPLEMENTATION_STATUS.md — Estado de implementación

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivitySlowing

ResponsivenessNo issues

Resources

GitHub Repository

Need Help?

Related Servers

Related MCP Connectors

QualeQuest — AI Agent Governance PlatformOAuth
com.qualequest
Multi-agent governance: task orchestration, compliance, decision validation, and ML predictions.
Autonomous Agents AI & Machine Learning
2
AgentMarketplace A
io.github.raditotev
MCP Server for an Agent Task Marketplace
Autonomous Agents Agent Orchestration
221
Frostbyte
dev.getfrostbyte
Roadmap, tasks, releases and user feedback your coding agent reads and writes over MCP.
Project Management Product Management Developer Tools
1
llm-orchestration-mcp-agent
ai.getvda
LLM Orchestration MCP Agent
Autonomous Agents Agent Orchestration AI & Machine Learning
1

Related MCP Servers

Rules MCP Server
Coding Agents Developer Tools
markacianfrani
F
license
A
quality
D
maintenance
An MCP server that allows coding agents to look up contextual rules and patterns on demand, providing just-in-time guidance for specific tasks like writing tests or authoring UI.
Updated a year ago (2025-07-25 11:19 UTC)
2
8
-
Spec Guard
Developer Tools Agent Orchestration Autonomous Agents
jpstone
A
license
Not graded
quality
D
maintenance
A methodology and MCP server for agent-driven software development where humans write specs and agents implement code, enforced by six mechanical gates to ensure spec validity, contracts, tests, and review.
Updated 3 months ago (2026-06-30 07:18 UTC)
17 npm
MIT
MCP Multi-Agent Server
Agent Orchestration Autonomous Agents Workplace & Productivity
sunnylabtv-crypto
A
license
Not graded
quality
C
maintenance
A multi-domain multi-agent system served over MCP (Model Context Protocol) with FastMCP, exposing specialized domain agents as tools to automate business workflows.
Updated a month ago (2026-08-05 05:58 UTC)
MIT
ArchNav MCP Server
Developer Tools Code Analysis Software Architecture
1020911401
F
license
Not graded
quality
C
maintenance
Architecture-aware code navigation system for AI agents, providing line-level symbol location, layered architecture detection, and dual-engine hybrid analysis with 22 MCP tools.
Updated 3 months ago (2026-06-30 10:32 UTC)
-