The Agent Skills Directory

[SAFE]: The skill provides legitimate industrial security monitoring functionality. All scripts perform local data analysis on network logs and packet captures without external network communication or suspicious file operations. The metadata discrepancy between the author field and license copyright (mahipal vs mukul975) appears to be a minor documentation inconsistency with no security implications.
[DATA_EXPOSURE]: No hardcoded credentials, sensitive file access, or network exfiltration patterns were identified. The network monitoring is passive and targets industrial protocol traffic (Port 502) as intended.
[EXTERNAL_DOWNLOADS]: The skill uses 'scapy', a well-known and trusted library for packet manipulation. No untrusted remote scripts or packages are downloaded.
[INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface for untrusted data (PCAP files and Zeek logs). While it lacks explicit sanitization or boundary markers for this data, its capabilities are strictly limited to local file reading and reporting, posing no risk of compromise or automated action execution.

detecting-modbus-protocol-anomalies