voip-sip-attack

Warn

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides an extensive collection of shell commands for network exploitation, such as SIP and RTP flooding for denial-of-service attacks (inviteflood, iaxflood, rtpflood) and network traffic interception using tcpdump and tshark.
  • [PRIVILEGE_ESCALATION]: Instructions include the use of sudo for high-impact tools like bettercap and commands to modify system-level kernel parameters, such as enabling IP forwarding (ip_forward) for man-in-the-middle attacks.
  • [DYNAMIC_EXECUTION]: Multiple files contain embedded scripts that generate and execute network traffic at runtime, including Python snippets for SIP INVITE flooding and RTP packet replaying, as well as Bash scripts for automating entire VoIP assessments.
  • [DATA_EXFILTRATION]: The skill documents methods for intercepting and decoding real-time voice communications (RTP) and performing offline cracking of SIP authentication digests using provided dictionary attack commands.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted network data, creating a surface for indirect prompt injection.
  • Ingestion points: Network traffic captures (PCAP files) parsed by tshark and tcpdump, and responses from remote SIP servers processed by tools like sipsak, svmap, and svwar as described in payloads.md and various guide files.
  • Boundary markers: None identified; instructions do not include delimiters or warnings to ignore embedded instructions within processed network data.
  • Capability inventory: The Bash tool provides broad capabilities including network communication, file system modification, and execution of the included Python and Bash scripts across all provided markdown documentation.
  • Sanitization: There is no evidence of sanitization, validation, or escaping of data retrieved from the network before it is passed to analysis tools or displayed to the agent.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 19, 2026, 04:36 PM