The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from network interfaces and capture files, creating a surface for potential indirect prompt injection attacks where malicious protocol traffic could attempt to influence the agent's analysis.
Ingestion points: Network traffic sniffing and pcap analysis via Scapy in SKILL.md; Zeek log parsing in scripts/agent.py.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present in the parsing logic.
Capability inventory: The skill's capabilities are limited to protocol analysis and printing reports to stdout; it lacks file-write or network-exfiltration functions.
Sanitization: The skill uses structural binary parsing (struct.unpack) and TSV field splitting, which provides structural validation of the input data.
[PRIVILEGE_ESCALATION]: The Modbus traffic monitoring functionality in SKILL.md utilizes Scapy's sniff function, which requires root or administrative privileges to access raw network sockets on most operating systems. While this is the standard and expected requirement for network monitoring tools, it grants the skill significant system visibility and access.

detecting-modbus-command-injection-attacks