tracking-threat-actor-infrastructure

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The content provides detailed, actionable dual-use instructions and code for discovering and pivoting across adversary infrastructure (searching for C2 servers, passive DNS/CT/WHOIS pivots, Shodan/Censys queries and fingerprinting), which could be used offensively despite defensive framing.

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill embeds and passes API keys directly in code and HTTP requests (e.g., "YOUR_SHODAN_API_KEY", headers with "APIKEY", and basic auth using api_key), which would require the LLM/agent to handle or insert secret values verbatim into generated output — an exfiltration risk.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). In scripts/process.py (and similarly scripts/agent.py), the runtime fetches and ingests outsider-authored free text/records from external sources like https://crt.sh/?q=%.{domain}&output=json (certificate transparency log entries) and parses their fields, and also queries passive DNS/WHOIS-style APIs using attacker-influenced --domain/--ip inputs.