The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides detailed architectural requirements for an 'implant' to execute actions upon receiving network triggers, specifically instructing it to 'Fork callback connection to callback_addr:callback_port'. This describes a remote command and control (C2) mechanism.
[DATA_EXFILTRATION]: The skill describes encoding sensitive 'callback' information (IP and Port) into covert network channels. While focusing on the trigger (ingress), the callback mechanism it facilitates is a primary method for data exfiltration.
[REMOTE_CODE_EXECUTION]: The skill provides the logical structure for a network-triggered remote access tool. It includes a specific payload structure containing callback instructions and cryptographic validation (SHA-1 key hash) to ensure only authorized attackers can trigger the dormant process.
[PROMPT_INJECTION]: The skill uses highly suspicious terminology including 'covert wake-up mechanisms', 'dormant processes', 'protocol steganography', and 'implant key', which are markers for malicious software development instructions.
[COMMAND_EXECUTION]: The skill requires the use of raw sockets (PF_PACKET, SOCK_RAW) and low-level packet crafting, which typically require elevated privileges (root/administrator) and are used for bypassing standard network stack security and logging.

covert-network-triggers