argo-ip-radar

python builtin, a high-precision, adaptive network diagnostic suite designed for low-bandwidth (500–600 kbps) and filtered environments. argo-ip-radar utilizes weighted subnet analysis and real-time performance feedback to identify the most stable Cloudflare entry points, specifically measuring WebSocket (WS) connection speeds for Argo Tunnel configurations.

---

Table of Contents

• Description
• Features
• Requirements
• Usage
• Algorithm
• License

---

Description

In regions with aggressive internet filtering and unstable connections, standard network scanners often fail due to fixed timeouts and high protocol overhead. argo-ip-radar is engineered specifically for these constraints. It treats each network subnet as a dynamic entity, “learning” which ranges provide the best WebSocket (WS) throughput and lowest latency through a floating-point weighting system.

By mimicking standard browser headers and using adaptive, median-based timeouts, the tool finds “clean” IPs that remain hidden from Deep Packet Inspection (DPI) while maximizing the utility of a limited 600 kbps link.

---

Features

• WebSocket Speed Verification Specifically tests the throughput of WS-compatible paths to ensure stability for 3x-ui and Argo tunnels.

• Adaptive Median Timeout Automatically adjusts connection limits based on current network jitter to prevent false block reports on unstable links.

• Weighted Subnet Discovery Uses a specialized algorithm to prioritize scanning in high-performing CIDR blocks based on historical success stored in subnets.json.

• High-Precision Throughput Measures speed as a floating-point value (kbps) to distinguish between marginal and stable connections without rounding errors.

• Memory-Buffered Logging Updates subnet performance scores in RAM and writes to disk only upon completion or interruption to protect network I/O performance.

---

Requirements

• Python 3.8+
• subnets.json Tracking file for Cloudflare CIDR blocks and their current performance weights.
• Network An active connection (optimized for 500–600 kbps bottlenecks).

---

Usage

1. Generate Weighted IPs

Run the generator to create an ips.txt file based on the current intelligence in subnets.json. Subnets with better historical performance are favored.

python ip-gen.py

---

2. Run the Radar

Execute the scanner to test WebSocket connection connectivity and true throughput.

python web-check.py

Note: Press Ctrl+C at any time to stop. The script will automatically save updated subnet weights to disk before exiting.

---

3. Sort Results

Filter the working IPs to identify the best candidates for your configuration.

python sorter.py

---

How it Works

The project utilizes an Exponentially Weighted Moving Average (EWMA) for its radar logic.

1. Exploration IPs are selected from multiple subnets based on their current weights.

2. Evaluation Each IP undergoes:

   • TCP connection
   • TLS handshake
   • Raw data transfer of /cdn-cgi/trace to simulate and measure WebSocket connection speed

3. Reinforcement

   • PASS: Subnet weight increases by (speed_kbps / 100.0)
   • BLOCK: Subnet weight is penalized by 10% (weight × 0.9)

4. Adaptation Future scans naturally prioritize high-performing subnets while avoiding filtered or throttled ranges.

---

License

This project is licensed under the MIT License.