skills/mukul975/anthropic-cybersecurity-skills/conducting-pass-the-ticket-attack/Gen Agent Trust Hub
conducting-pass-the-ticket-attack
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides detailed workflows and automated command generation for extracting Kerberos tickets and performing lateral movement using tools like Mimikatz, Rubeus, and Impacket. These techniques are standard for red-teaming and are provided as the skill's primary educational purpose.
- [PROMPT_INJECTION]: The scripts agent.py and process.py ingest external data from Windows logs and Kerberos ticket files. This creates a surface for indirect prompt injection where malicious data could attempt to influence the agent's interpretation of the attack or detection results.
- Ingestion points: agent.py reads JSON logs; process.py reads .kirbi and .ccache binary files.
- Boundary markers: No boundary markers or specific 'ignore' instructions are used when processing external data.
- Capability inventory: The scripts perform data parsing, extraction, and command generation; they do not automatically execute system commands based on the ingested data.
- Sanitization: The scripts rely on standard library parsers (json, struct) but do not implement specific sanitization or validation for content passed to the user or an LLM.
Audit Metadata