darkweb-intel
Audited by Socket on Sep 19, 2026
5 alerts found:
SecurityAnomalyx4SUSPICIOUS: the skill is internally coherent for dark-web OSINT, but its actual footprint is a high-risk offensive/recon capability for an AI agent. It enables Tor-routed investigation, onion enumeration, marketplace monitoring, and actor profiling, plus evasion/OPSEC techniques and crypto-related workflow guidance. There is no clear credential theft or deceptive exfiltration, so this is not confirmed malware, but it materially increases security risk and operational abuse potential.
This fragment is an operational dark-web investigation playbook, not executable package malware. It contains network collection and monitoring commands, local evidence handling, and destructive cleanup procedures. The principal risks are unsafe secret handling, an unprotected generated PGP key, metadata disclosure to external services, shell-script correctness issues, and potentially destructive deletion commands. No clear malicious payload, unauthorized persistence, credential theft, or covert exfiltration is evident in the shown portion. Because the file is truncated, this assessment applies only to the visible content and cannot establish the behavior of omitted sections.
The fragment does not show clear malware or an intentional supply-chain backdoor. It documents dual-use overlay-network enumeration and sensor deployment. The main security concerns are operational: local credential handling, mutable `latest` images, runtime package installation, potentially exposed administration ports, and especially Docker `--privileged` for Lokinet. The code should not be run unchanged in production, but the supplied content provides insufficient evidence of malicious intent or data exfiltration.
This fragment is an operational dark-web intelligence and threat-actor monitoring guide, not ordinary npm runtime code. It shows extensive network collection and local evidence processing, but no clear malware payload, data theft from the host, reverse shell, cryptomining, or destructive system modification. The main risks are potentially unauthorized surveillance, leakage of search terms or API credentials, arbitrary network access through shell interpolation, cron-based monitoring persistence, and anti-forensic cleanup commands. Malware intent is low based on the visible content, while operational security and misuse risk are moderate. Review of the omitted portions would still be required before assessing the complete file.
No direct malware or covert supply-chain attack is evident because the input is documentation rather than executable package code. The procedures present moderate security and privacy risk if copied literally: API keys and sensitive identifiers may leak, breach-password data may be logged, raw responses may be sent to third parties, and unsafe placeholder interpolation may enable command or request injection. The material should be treated as an operational playbook requiring strict authorization, secret management, shell-safe parameter handling, redaction, least-privilege logging, secure evidence retention, and isolated execution.