# Flowlines MCP server

> Analyze MCP server usage, investigate sessions, and instrument MCP servers for observability.

- Repository: https://github.com/flowlines-ai/plugins
- Page: https://hostedmcp.org/servers/flowlines-ai/plugins
- Transport: remote
- Stars: 0
- License: MIT
- Latest version: 2.0.0

## Install

### Claude Code

```sh
claude mcp add --transport http plugins https://api.flowlines.ai/mcp
```

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

```json
{
  "mcpServers": {
    "plugins": {
      "url": "https://api.flowlines.ai/mcp"
    }
  }
}
```

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

```json
{
  "servers": {
    "plugins": {
      "type": "http",
      "url": "https://api.flowlines.ai/mcp"
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"plugins","type":"http","url":"https://api.flowlines.ai/mcp"}'
```

### Codex

```sh
codex mcp add plugins --url https://api.flowlines.ai/mcp
```

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

```toml
[mcp_servers.plugins]
url = "https://api.flowlines.ai/mcp"
```

### Gemini CLI

```sh
gemini mcp add -s user --transport http plugins https://api.flowlines.ai/mcp
```

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

```json
{
  "mcpServers": {
    "plugins": {
      "url": "https://api.flowlines.ai/mcp",
      "type": "http"
    }
  }
}
```

## FAQ

### What does the Flowlines MCP server do?

It connects your MCP client to a Flowlines workspace to inspect how people use your MCP servers, track changes since releases, and identify session failures.

### How can I instrument my MCP server with Flowlines?

You can ask your MCP client to onboard you to Flowlines, which calls the onboarding tool to generate an instrumentation plan and verifies the telemetry.

### Is my data private when using Flowlines?

The server reads production sessions and tool calls, so all returned data should be treated as confidential. You can control what is recorded via the get-started page.

### What should I do if I encounter MCP connection errors?

You can use the `flowlines-doctor` skill to diagnose missing or incomplete data and check client status or local logs.
