Skip to main content

MCP Tools

The Model Context Protocol (MCP) enables AI agents to access external tools and integrations. Strongly AI ships with 136 pre-built MCP servers covering databases, APIs, productivity tools, and more.

Architecture​

Single MCP App per org (one K8s pod)
  ├── Tool Registry (in-memory registered functions)
  ├── 136 MCP servers (each in its own mcps/ folder)
  ├── /mcp endpoint (standard MCP protocol)
  └── MongoDB persistence (survives pod restarts)

When a user enables an MCP:

  1. Tool definitions loaded into the in-memory registry
  2. User config (API keys) stored securely in MongoDB
  3. Tools appear in the aggregated /mcp tools/list response
  4. Workflow agents can call tools via tools/call

Available MCP Servers​

Search & Web​

  • Brave Search — web, image, video, news search
  • DuckDuckGo — privacy-focused search (no API key)
  • Exa — AI-powered semantic search
  • Firecrawl — web scraping and data extraction
  • Google Translate — translation for 100+ languages

Development & DevOps​

  • GitHub — repos, issues, PRs, code search (26 tools)
  • Docker Hub — image search, tags, repositories
  • GitLab — repository management
  • SonarQube — code quality analysis
  • Postman — API testing and management
  • Netlify / Render — deployment platforms

Databases​

  • Redis — key-value, hashes, lists, sets
  • Elasticsearch — full-text search and analytics
  • MongoDB — document CRUD, aggregation
  • Neo4j — graph database Cypher queries
  • SingleStore — distributed SQL
  • Couchbase — NoSQL document database
  • CockroachDB — distributed SQL

Productivity​

  • Atlassian — Jira, Confluence, Bitbucket (16 tools)
  • Notion — pages, databases, search
  • Todoist — task management
  • Google Tasks — task lists

Communication​

  • Mailgun — transactional email
  • LinkedIn — post content, profile
  • Webex / Mattermost / RocketChat — messaging

Finance & Commerce​

  • Stripe — payments, customers, subscriptions
  • Coinbase — cryptocurrency data
  • Razorpay — payment processing

AI & Data​

  • Wolfram Alpha — computational knowledge
  • DeepL — neural machine translation
  • ElevenLabs — text-to-speech
  • OpenWeather — weather data and forecasts

Monitoring​

  • Grafana — dashboards, alerts, data sources
  • Dynatrace — APM monitoring

Utilities​

  • HTTP Fetch — generic web requests (no API key)
  • RSS — feed parsing (no API key)
  • Time — timezone operations

...and 100+ more. See the full catalog at /workflow-tools.

Configuring an MCP​

  1. Navigate to Workflows > Tools
  2. Find the MCP server you want to use
  3. Click Details and enter the required configuration (API keys, URLs)
  4. Toggle the MCP On — tools become available immediately
  5. Use in workflows via the MCP Tools Provider node

REST API​

# Register an MCP with config
POST /api/v1/mcps/register
{
  "mcp_id": "mcp-brave-search",
  "mcp_name": "Brave Search",
  "tools": [...],
  "config": { "braveApiKey": "BSAxxxx" },
  "enabled": true
}

# Toggle MCP on/off
PATCH /api/v1/mcps/{mcp_id}/toggle
{ "enabled": true }

# List all MCPs
GET /api/v1/mcps

# List all tools (aggregated)
GET /api/v1/mcps/tools

# MCP Protocol (standard)
POST /mcp
{ "jsonrpc": "2.0", "method": "tools/list", "id": 1 }

Adding Custom MCPs​

Each MCP server is a Python module in apps/mcp-gateway/app/mcps/{name}/:

# mcps/my_custom_tool/__init__.py
import httpx
from typing import Dict, Any

TOOLS = [
    {"name": "my_tool", "description": "Does something useful",
     "inputSchema": {"type": "object", "properties": {"query": {"type": "string"}}, "required": ["query"]}},
]

async def execute(tool_name: str, arguments: Dict[str, Any], config: Dict[str, str]) -> Dict[str, Any]:
    async with httpx.AsyncClient() as c:
        r = await c.get("https://api.example.com/search", params={"q": arguments["query"]})
        return {"content": [{"type": "text", "text": r.text}]}