# Conduct AI MCP server

> Manage AI agents, trigger workflows, and enforce team security policies across AI clients like Claude Code and Cursor.

- Repository: https://github.com/sseshachala/conduct-cli
- Page: https://hostedmcp.org/servers/sseshachala/conduct-cli
- Transport: stdio
- Stars: 0
- Latest version: 0.6.10

## Install

### Claude Code

```sh
claude mcp add conduct-cli -- uvx conduct-cli
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "conduct-cli": {
      "command": "uvx",
      "args": [
        "conduct-cli"
      ]
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "conduct-cli": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "conduct-cli"
      ]
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"conduct-cli","type":"stdio","command":"uvx","args":["conduct-cli"]}'
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "conduct-cli": {
      "command": "uvx",
      "args": [
        "conduct-cli"
      ]
    }
  }
}
```

### Codex

```sh
codex mcp add conduct-cli -- uvx conduct-cli
```

### Codex config.toml (`~/.codex/config.toml`)

```toml
[mcp_servers.conduct-cli]
command = "uvx"
args = ["conduct-cli"]
```

### Gemini CLI

```sh
gemini mcp add -s user conduct-cli uvx -- conduct-cli
```

### Gemini settings.json (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "conduct-cli": {
      "command": "uvx",
      "args": [
        "conduct-cli"
      ]
    }
  }
}
```

## FAQ

### How do I install Conduct AI?

You can install the official CLI using pip with the command `pip install conduct-cli`.

### What is ConductGuard?

ConductGuard is an AI tool fleet management system that allows security teams to set policies that are automatically enforced across developer AI sessions.

### Can I use Conduct AI in my CI/CD pipeline?

Yes, the `conduct test --all` command returns an exit code of 0 if all pass and 1 if any fail, making it suitable for CI environments.
