# agent-observability MCP server

> Trace AI agent execution by monitoring tool calls, errors, token usage, and costs through a local-first dashboard.

- Repository: https://github.com/KryptosAI/agent-observability
- Page: https://hostedmcp.org/servers/kryptosai/agent-observability
- Transport: stdio
- Stars: 2
- License: MIT
- Latest version: 1.0.7

## Install

### Claude Code

```sh
claude mcp add agent-observability -- npx -y agent-obs
```

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

```json
{
  "mcpServers": {
    "agent-observability": {
      "command": "npx",
      "args": [
        "-y",
        "agent-obs"
      ]
    }
  }
}
```

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

```json
{
  "servers": {
    "agent-observability": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "agent-obs"
      ]
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"agent-observability","type":"stdio","command":"npx","args":["-y","agent-obs"]}'
```

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

```json
{
  "mcpServers": {
    "agent-observability": {
      "command": "npx",
      "args": [
        "-y",
        "agent-obs"
      ]
    }
  }
}
```

### Codex

```sh
codex mcp add agent-observability -- npx -y agent-obs
```

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

```toml
[mcp_servers.agent-observability]
command = "npx"
args = ["-y", "agent-obs"]
```

### Gemini CLI

```sh
gemini mcp add -s user agent-observability npx -- -y agent-obs
```

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

```json
{
  "mcpServers": {
    "agent-observability": {
      "command": "npx",
      "args": [
        "-y",
        "agent-obs"
      ]
    }
  }
}
```

## Tools (6)

- `obs_record_tool_call`
- `obs_record_token_usage`
- `obs_record_decision`
- `obs_record_grade`
- `obs_get_session_report`
- `obs_list_sessions`

## FAQ

### Does my data leave my machine?

No, all data is stored in a local SQLite database and no data leaves your machine.

### How can I see my agent's sessions?

You can view your sessions by running the dashboard command, which hosts a local web interface.

### What happens if my agent cannot self-report?

You can use Proxy mode to wrap any MCP server, which automatically traces every tool invocation.
