The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/agent.py script executes the tshark binary via subprocess.run(). This is documented as the skill's primary function. The implementation is secure as it uses list-based arguments instead of a shell string, preventing command injection via the filename or other parameters.\n- [PROMPT_INJECTION]: The skill represents an indirect prompt injection surface because it parses untrusted network traffic (PCAP files) and extracts human-readable strings into its output. If a malicious packet contains instructions designed to manipulate an LLM, those instructions could be processed by the agent when it reviews the analysis results.\n
Ingestion points: Network packet data is ingested from PCAP/PCAPNG files via TShark extraction commands in scripts/agent.py.\n
Boundary markers: The skill does not use specific delimiters or instructions to warn the agent that the extracted DNS queries, hostnames, and URLs should be treated as untrusted data.\n
Capability inventory: The skill possesses the capability to execute shell commands (tshark) and write JSON reports to the local filesystem.\n
Sanitization: The script does not perform sanitization of extracted strings (such as URL paths or DNS query names) to remove potential prompt injection sequences before they are included in the report.

performing-network-traffic-analysis-with-tshark