offensive-data-exfiltration

Fail

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: CRITICALDATA_EXFILTRATIONCREDENTIALS_UNSAFEPERSISTENCEOBFUSCATIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill provides comprehensive methodologies and functional code for exfiltrating sensitive information through multiple egress channels.
  • Implements DNS exfiltration techniques using TXT and CNAME records, including 'slow-drip' methods designed to evade entropy-based detection.
  • Provides instructions for HTTPS tunneling and domain fronting to mask data theft as legitimate web traffic to CDNs.
  • Contains functional code for ICMP tunneling, allowing data to be hidden inside ping packets.
  • Facilitates exfiltration to cloud storage providers (AWS S3, Azure, GCS) using presigned URLs and SAS tokens that require no authentication on the target host.
  • Abuses legitimate services such as Slack webhooks, GitHub Gists, and Pastebin for data transit.
  • Details an email-based 'draft method' where data is saved in account drafts to avoid generating network traffic associated with sent messages.
  • [CREDENTIALS_UNSAFE]: The skill instructions specifically target the collection of high-value system credentials.
  • Includes explicit commands to access and download the /etc/shadow file containing system password hashes.
  • Directs the agent to search for and exfiltrate private SSH keys (id_rsa) from home directories.
  • [PERSISTENCE]: Provides mechanisms to maintain long-term access and ensure recurring data collection.
  • Includes Linux cron job configurations to execute exfiltration scripts on a set schedule during business hours to blend with normal traffic.
  • Provides PowerShell commands to register hidden Windows Scheduled Tasks that trigger exfiltration activities every 30 minutes.
  • [OBFUSCATION]: Details multiple techniques to conceal data and bypass security inspection.
  • Implements steganography to hide encrypted data within images (LSB), audio (WAV), and PDF metadata fields.
  • Provides implementation for network timing channels, which encode data in the duration of delays between transmissions to remain undetectable by traditional traffic analysis.
  • Includes instructions for HTTP header encoding to hide exfiltrated data within standard protocol headers.
  • [EXTERNAL_DOWNLOADS]: References and provides links to numerous third-party offensive tools hosted on external repositories, including dnscat2, iodine, chisel, and ptunnel-ng.
  • [COMMAND_EXECUTION]: Instructs on the use of complex shell and PowerShell pipelines for data staging, including mass file collection, local encryption using AES-256-GCM, and splitting archives into small, stealthy chunks.
  • [DYNAMIC_EXECUTION]: Contains Python scripts that perform raw socket manipulation for ICMP exfiltration and use the Scapy library to inject data into IP header fields at runtime.
Recommendations
  • CRITICAL: 1 infected file(s) detected - DO NOT USE
  • AI detected serious security threats
Audit Metadata
Risk Level
CRITICAL
Analyzed
Aug 27, 2026, 02:27 PM
Security Audit — agent-trust-hub — offensive-data-exfiltration