# inspect-logs-mcp

MCP server for exploring [inspect_ai](https://github.com/UKGovernmentBEIS/inspect_ai) evaluation logs from Claude Code.

## What it does

This MCP server gives Claude Code, Cursor, and other LLMs direct access to your inspect_ai evaluation logs, allowing you to:

- **List logs**: See all your evaluation runs with metadata
- **View summaries**: Get detailed evaluation results, scores, and token usage
- **Inspect samples**: View full conversation history for any sample
- **Search & filter**: Find logs by task, model, date, or status
- **Compare runs**: Side-by-side comparison of two evaluation runs
- **Aggregate stats**: Get statistics across multiple runs

## Installation

Currently only source installation is supported.

```bash
git clone https://github.com/PranshuSrivastava/inspect-logs-mcp.git
cd inspect-logs-mcp
pip install -e .
```

## Configuration

To use with Claude code in your current directory, run the following command:

```bash
claude mcp add --transport stdio inspect-logs inspect-logs-mcp
```

To use with Claude code in your global configuration, run the following command:

```bash
claude mcp add --scope user --transport stdio inspect-logs inspect-logs-mcp
```

To use with Cursor, Antigravity or other IDEs, add the following to your mcp config file:

```json
{
    "mcpServers": {
        "inspect-logs": {
            "command": "inspect-logs-mcp",
            "env": {
                "INSPECT_LOG_DIR": "./logs"
            }
        }
    }
}
```

## Environment Variables

| Variable | Description | Default |
|----------|-------------|---------|
| `INSPECT_LOG_DIR` | Directory containing `.eval` log files | `./logs` |
| `INSPECT_LOGS_MCP_DEFAULT_LIMIT` | Default number of logs limit | `50` |
| `INSPECT_LOGS_MCP_MAX_LIMIT` | Maximum number of logs limit | `500` |

## How it Works

The server reads `.eval` files (which are ZIP archives containing JSON) using the `inspect_ai.log` API. All file operations happen in memory - no files are extracted to disk, so your logs directory stays clean.

## Requirements

- Python 3.10+
- inspect-ai >= 0.3.70
- mcp >= 1.0.0