raven-monitor 0.6.0

Cross-platform system monitor with TUI, web dashboard, and remote monitoring

Raven 🐦‍⬛

A system monitor that works on Linux, BSD, macOS, and Windows.

---

🌟 Key Features

• ⚡ Gorgeous Dashboards:
  • TUI Dashboard: Built with Textual, offering smooth animations, color themes, real-time widgets, and flicker-free updates.
  • Web Dashboard: An elegant web UI powered by FastAPI and Vanilla HTML/CSS/JS. Features premium loading shimmers, connection status toasts, light/dark theme persistence (to prevent FOUC), and dynamic charts.
• 🐦‍⬛ Neofetch-style Fetch: A quick terminal command to fetch system hardware specs, platform info, and resources.
• 🔒 Remote Monitoring: Run a secure agent on a remote server with timing-attack resistant API key authentication, and visualize its metrics locally.
• 🔌 Extensible Plugins: Easily write your own metrics collectors by inheriting from MonitorPlugin.
• 📈 High Performance: Metrics are queried in parallel via thread pools, with smart caching and lightweight process queries to minimize CPU overhead.
• 📥 Multi-format Exporting: Print metrics directly to your terminal or save them as CSV, JSON, or plaintext.

---

🛠 Project Structure

Raven is structured cleanly to separate collection logic, plugins, and frontends:

• raven/cli.py: CLI routing and argument definitions.
• raven/config.py: TOML configuration schemas and verification logic.
• raven/core/: Central coordinator containing the collector agent, type definitions, background server runner, and standard protocol interfaces.
• raven/plugins/: Discovered monitoring plugins (CPU, Memory, Disk, Containers, Network, Sensors, Processes, etc.).
• raven/tui/: Textual terminal widgets and layout CSS stylesheets (.tcss).
• raven/web/: FastAPI backend and static web dashboard assets.
• raven/remote/: Server agent and client tools for remote metric sync.
• raven/export/: Formatted data exporters.

---

🚀 Getting Started

1. Prerequisites

• Python 3.11+
• uv (recommended) or pip

2. Installation & Setup

To install dependencies and prepare the environment:

# Clone the repository
git clone https://github.com/salvatorecorvaglia/raven.git
cd raven

# Sync virtual environment and download dependencies
uv sync --all-extras --dev

---

💻 Usage & CLI Guide

Raven offers a simple command-line structure routed through raven/cli.py:

1. Start the TUI (Default)

Run Raven with no arguments to launch the Textual TUI dashboard:

uv run raven

2. Start the Web Dashboard

Expose a web-based dashboard utilizing FastAPI:

uv run raven web
# Options:
#   --host Hostname to bind to (e.g. 127.0.0.1)
#   --port Port to run server on (e.g. 8080)

3. Start a Remote Monitoring Agent

To monitor a remote server, run the daemon agent on the remote host:

uv run raven serve --host 127.0.0.1 --port 9090

4. Connect to a Remote Agent

You can direct your TUI, Web server, or exporters to read metrics from a running remote agent:

uv run raven --remote http://192.168.1.50:9090

5. Fetch a Quick System Summary

Get a quick neofetch-style output of your server specifications:

uv run raven fetch

6. Export Metrics to Stdout

Print system details to the terminal in JSON, CSV, or formatted text:

# Print all modules once as text
uv run raven print

# Print specific modules formatted as JSON
uv run raven print cpu memory --format json

---

⚙️ Configuration

Raven searches for settings in the following order:

1. --config / -c CLI option.
2. ./raven.toml in the current working directory.
3. ~/.config/raven/raven.toml.
4. Fall back to internal defaults.

Refer to raven.example.toml for customising ports, intervals, enabled modules, and security parameters:

[general]
refresh_interval = 2        # seconds between updates
theme = "dark"

[modules]
cpu = true
memory = true
disk = true
network = true
processes = true
users = true
sensors = true
containers = true

[web]
enabled = false
host = "127.0.0.1"
port = 8080
api_key = ""                # Empty = no authentication

[remote]
enabled = false
host = "127.0.0.1"
port = 9090
api_key = ""

---

🔌 Creating Custom Plugins

All metrics collection modules are structured as plugins. To add your own custom collector:

1. Create a Python file under the raven/plugins/ directory.
2. Subclass MonitorPlugin and implement the abstract methods:

from raven.plugins.base import MonitorPlugin

class CustomMetricsPlugin(MonitorPlugin):
    name = "my_custom_metrics"
    category = "general"

    def is_available(self) -> bool:
        # Check system compatibility or dependencies here
        return True

    def collect(self) -> dict:
        # Collect and return your metrics
        return {
            "custom_metric_1": 42,
            "status": "online"
        }

3. Dynamic Discovery & Loading: Built-in and custom plugins are dynamically discovered from the raven/plugins/ directory using pkgutil.iter_modules. Your custom plugin will be loaded automatically if is_available() returns True. You do not need to register custom modules in raven.toml unless they require custom options/configurations passed to __init__.

[!TIP] Configuration Support: If your plugin requires custom settings or needs to read the global configuration, define an __init__(self, config=None) constructor. The plugin manager checks the constructor signature and automatically passes the RavenConfig instance to it.

[!TIP] Thread Safety: If your plugin keeps state between collection cycles (e.g. counters or cached query data), use a thread lock (such as threading.Lock) to make sure it is safe to access from parallel collection threads.

---

🧪 Testing & Development

We use pytest for automated test suites. Before submitting pull requests, run:

# Run the test suite
uv run pytest

# Check code formatting & lint issues
uv run ruff check

# Apply formatting
uv run ruff format

---

🤝 Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

🔐 Security & Authentication

Raven supports timing-attack resistant API key authentication to secure web and remote server execution:

• Configuring API Keys: Set the api_key under the [web] and [remote] headers in your raven.toml. Leaving the value empty disables authentication.
• Timing-Safe Checks: Server-side comparisons use hmac.compare_digest to safeguard against timing-attacks.
• REST & WebSockets API:
  • REST endpoints require client requests to supply the X-API-Key HTTP header.
  • WebSocket streams authenticate via the first message sent over the socket to avoid leaking credentials in URL parameters. Connection failure closes the socket with status code 4001.
• Obfuscated Browser Storage: The Web Dashboard automatically obfuscates keys saved in localStorage and sessionStorage using a XOR cipher with a secure salt, preventing raw clear-text exposure.
• Clean URLs: Query parameters containing keys (e.g., ?api_key=...) are automatically read, stored in obfuscated format, and stripped from the browser's address bar to keep query logs clean.
• Open Bind Warnings: Starting a server on 0.0.0.0 (all network interfaces) without a configured API key prints a console warning advising either key configuration or binding to 127.0.0.1.
• Referrer Privacy: To prevent credential and path leakage, all API responses include a Referrer-Policy: no-referrer header, and the Web Dashboard includes a <meta name="referrer" content="no-referrer"> directive.

If you discover a security vulnerability, please see our Security Policy.

📝 License

Distributed under the MIT License. See LICENSE for more information.

---

Author: Salvatore Corvaglia