The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation references well-known Python libraries including scapy and pymodbus. These are standard industry tools for network packet manipulation and industrial protocol communication, and their inclusion is appropriate for the skill's stated purpose of OT security monitoring.
[INDIRECT_PROMPT_INJECTION]: The provided scripts (SKILL.md body and scripts/agent.py) ingest external data sources such as network capture files (PCAPs) and Zeek transaction logs. While this constitutes an attack surface for indirect prompt injection, the scripts utilize structured parsing (via scapy or tab-separated log parsing) and standard type conversion, which limits the potential for adversarial data to influence the agent's core logic or instructions beyond the analysis results.
[COMMAND_EXECUTION]: The Python scripts perform data processing and parsing operations using standard libraries like struct and collections. No instances of arbitrary shell command execution, subprocess spawning with untrusted input, or dynamic code evaluation (e.g., eval(), exec()) were identified.

detecting-modbus-protocol-anomalies