The Agent Skills Directory

[SAFE]: The skill is a legitimate security utility designed for DNS tunneling detection. It analyzes local PCAP files for statistical anomalies using the well-known scapy library.\n- [COMMAND_EXECUTION]: The Python script scripts/agent.py processes packet data and generates a structured JSON report. No dangerous command patterns, privilege escalation attempts, or unauthorized network operations were found in the code.\n- [DATA_EXFILTRATION]: No unauthorized data transmission or hardcoded credentials were detected. The script is restricted to reading local files and writing analysis results to a local JSON file.\n- [PROMPT_INJECTION]: The skill ingests untrusted DNS query names from PCAP files (Ingestion point: scripts/agent.py). Boundary markers are absent. The script reads files and writes results (Capabilities: scapy.rdpcap, json.dump). No specific sanitization of query names is performed for LLM context. This is a common surface for indirect prompt injection but is handled safely within the script's intended use case.

performing-dns-tunneling-detection