@sequa-ai/sequa-mcp v1.0.3

Sequa MCP

The missing brain for your AI dev tools

🤔 What is Sequa?

Sequa is a Contextual Knowledge Engine that unifies code, documentation and tickets across multiple repositories and streams that live context to any LLM‑powered assistant. By giving tools like Cursor or Claude deep, always‑current project knowledge, Sequa helps them answer architecture‑level questions, generate more accurate code and slash hallucinations.

🖥️ What is Sequa MCP?

sequa‑mcp is a tiny proxy that lets any STDIO‑based AI client talk to your Sequa workspace using the Model Context Protocol (MCP). It forwards STDIO traffic to Sequa’s streamable HTTP MCP endpoint - so IDEs that only support the command transport can connect with zero extra infrastructure.

Why not just use a URL?

Most IDEs currently speak MCP over STDIO commands and assume the proxy is responsible for networking. Sequa exposes an advanced bidirectional HTTP stream, not SSE, so direct url: configs will not work yet. Until IDEs add first‑class support, always configure Sequa through the command/args option shown below.

🚀 Quick Start

Via NPX

npx -y @sequa-ai/sequa-mcp@latest  https://mcp.sequa.ai/<endpoint>

Via Docker

docker run -i --rm --network host sequa/sequa-mcp:latest  https://mcp.sequa.ai/<endpoint>

🔌 Connect Your Favourite Tools

Replace https://mcp.sequa.ai/<endpoint> with your actual Sequa MCP URL. Always use the command style until IDEs support HTTP‑stream URLs directly.

Cursor

~/.cursor/mcp.json

{
  "mcpServers": {
    "sequa": {
      "command": "npx",
      "args": [
        "-y",
        "@sequa-ai/sequa-mcp@latest",
        " https://mcp.sequa.ai/<endpoint>"
      ]
    }
  }
}

Claude Desktop

Settings ➜ Developer ➜ Edit Config

{
  "mcpServers": {
    "sequa": {
      "command": "npx",
      "args": [
        "-y",
        "@sequa-ai/sequa-mcp@latest",
        "https://mcp.sequa.ai/<endpoint>"
      ]
    }
  }
}

Windsurf

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "sequa": {
      "command": "npx",
      "args": [
        "-y",
        "@sequa-ai/sequa-mcp@latest",
        "https://mcp.sequa.ai/<endpoint>"
      ]
    }
  }
}

VS Code

.vscode/mcp.json

{
  "servers": {
    "sequa": {
      "command": "npx",
      "args": [
        "-y",
        "@sequa-ai/sequa-mcp@latest",
        "https://mcp.sequa.ai/<endpoint>"
      ]
    }
  }
}

Cline (Claude Dev‑Tools)

~/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json

{
  "mcpServers": {
    "sequa": {
      "command": "npx",
      "args": [
        "-y",
        "@sequa-ai/sequa-mcp@latest",
        "https://mcp.sequa.ai/<endpoint>"
      ],
      "disabled": false,
      "autoApprove": []
    }
  }
}

Highlight AI

1. Click the plugins icon (@) ➜ Installed Plugins ➜ Custom Plugin ➜ Add using a command

2. Use:

   npx -y @sequa-ai/sequa-mcp@latest https://mcp.sequa.ai/<endpoint>

Augment Code

npx @sequa-ai/sequa-mcp@latest https://mcp.sequa.ai/<endpoint>

Or in augment_config.json:

{
  "mcpServers": {
    "sequa": {
      "command": "npx",
      "args": [
        "-y",
        "@sequa-ai/sequa-mcp@latest",
        "https://mcp.sequa.ai/<endpoint>"
      ]
    }
  }
}

⚙️ How It Works

IDE / Agent ⇄ Sequa MCP (local proxy) ⇄ Sequa Workspace (HTTP‑stream MCP)

1. Your IDE writes MCP requests on STDIO.
2. sequa‑mcp streams them over HTTPS to the Sequa workspace.
3. Sequa enriches the requests with real‑time, multi‑repo context and streams back partial results.
4. The proxy pipes the bytes straight to your IDE for instant feedback.