Skip to content
Conduit

mcp server

AIOpStack

Describe how to archieve human in loop to approve mcp call using langgraph

hyperaiops/AIOpStackcommunitylistedunreviewed
Use tool

Set up

Runs on your computer. Pick your agent and operating system.

Before you start

  • uv (`curl -LsSf https://astral.sh/uv/install.sh | sh`)
~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "aiopstack": {
      "command": "uvx",
      "args": [
        "aiopstack"
      ]
    }
  }
}

same format Cursor, GitHub Copilot CLI, Junie, JetBrains AI Assistant, Augment (Auggie CLI), Factory Droid, Kiro, Devin CLI +12

Values in angle brackets are placeholders. Restart Claude Desktop after editing its config.Claude Desktop docs

README

synced Oct 23, 2026

AIOpStack

Python ≥3.8 LangChain v0.3.25 LangGraph v0.4.8 LangChain MCP Adapters v0.1.7

Build your AIOps MCP Agent within 5 minutes.

AIOpStack is a collection of AI Operational Agents built with Langchain/LangGraph and Streamlit‑based GUI for operational interaction and visualization.

🌐 English | 简体中文

🎯 Motivation

  1. Fast‑boot – AIOpStack drastically reduces study and setup time for DevOps, Sysadmins and Developers.
  2. Lightweight – IDEs like Cursor and VSCode with Cline are resource‑heavy. AIOpStack stays minimal and nimble.
  3. Local Deployment – Deploy locally to access private environments.
  4. Free & Open – Fully open‑source, no vendor lock‑in or licensing fees.

🚀 Features

  • OpenAI‑compatible LLM API Integration – Connect to any OpenAI‑compatible LLM endpoint.
  • MCP Integration – Seamless bridge between LLMs and popular MCP tools (e.g., Kubernetes, Ansible).
  • Human‑in‑the‑loop Feedback – Pause for confirmation or iterative refinement at key steps.
  • Pure Python & GUI‑free – Fully Python‑powered: no frontend skills required for reuse or extension.

Video Demo

Agents Repository

Agent NameDescriptionReference Link
langchain-mcp-adapters-chatbotA chatbot application that leverages langchain-mcp-adapters, fully inheriting the configuration from the official Langchain project. Useful for testing MCP servers.langchain-mcp-adapters

📖 Quick Start

Requirements

  • Python 3.8+
  • OpenAI‑compatible LLM API URL & Key

Installation

pip install aiopstack

Usage Examples

run aiops with no parameter

aiops

the project can be listening at localhost:8501 by default.

⚙️ MCP Settings

Note:

The settings must be written in JSON format.

  • For stdio mode:
    • The args parameter must use an absolute path.
    • Windows uses \\ for paths, Linux uses /.
    • Ensure all required Python dependencies are pre-installed.
  • For sse and streamable_http modes:
    • It is recommended to use a fully qualified domain name (FQDN) or a valid IP address to avoid connectivity issues.
MCP Configuration Example

The following configurations correspond to test examples located in the test_mcp_servers/ directory.

{
  "math": {
    "command": "python",
    "args": ["C:\\Users\\yfxue\\PycharmProjects\\aiopstack\\test_mcp_server\\math_server.py"],
    "transport": "stdio"
  },
  "weather": {
    "url": "http://192.168.2.103:8000/sse/",
    "transport": "sse"
  }
}

Strongly recommended: follow up the official document langchain-mcp-adapters

🤝 Contributing

We welcome contributions! If you have suggestions or feature requests, feel free to open an issue or submit a pull request.

Development Setup

  1. Clone the repository:
git clone https://github.com/hyperaiops/AIOpStack.git
  1. Install dependencies:
cd src/aiopstack
pip install -r requirements.txt
  1. Run Streamlit
streamlit run app.py
↑↓ navigate openesc close