# CodeImpact MCP server

> Analyzes TypeScript and JavaScript dependency graphs to provide risk scores and blast-radius assessments for local code changes.

- Repository: https://github.com/vk0dev/code-impact-mcp
- Page: https://hostedmcp.org/servers/vk0dev/code-impact-mcp
- Transport: stdio
- Stars: 1
- License: MIT
- Latest version: 1.6.10

## Install

### Claude Code

```sh
claude mcp add code-impact-mcp -- npx -y @vk0/code-impact-mcp
```

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

```json
{
  "mcpServers": {
    "code-impact-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@vk0/code-impact-mcp"
      ]
    }
  }
}
```

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

```json
{
  "servers": {
    "code-impact-mcp": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@vk0/code-impact-mcp"
      ]
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"code-impact-mcp","type":"stdio","command":"npx","args":["-y","@vk0/code-impact-mcp"]}'
```

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

```json
{
  "mcpServers": {
    "code-impact-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@vk0/code-impact-mcp"
      ]
    }
  }
}
```

### Codex

```sh
codex mcp add code-impact-mcp -- npx -y @vk0/code-impact-mcp
```

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

```toml
[mcp_servers.code-impact-mcp]
command = "npx"
args = ["-y", "@vk0/code-impact-mcp"]
```

### Gemini CLI

```sh
gemini mcp add -s user code-impact-mcp npx -- -y @vk0/code-impact-mcp
```

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

```json
{
  "mcpServers": {
    "code-impact-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@vk0/code-impact-mcp"
      ]
    }
  }
}
```

## FAQ

### What languages does it support?

It provides full TypeScript and JavaScript support, while Python support is limited to file-level path analysis.

### Does this require a database or cloud setup?

No, it is a local-first tool that builds an in-memory dependency graph with no database or cloud requirement.

### How is the risk score calculated?

The risk score is a value between 0 and 1, calculated by dividing the number of affected files by the total number of files.

### Can it be used as a pre-commit hook?

Yes, it includes a Husky-only helper to wire the gate runner into your pre-commit workflow.