{
  "id": "c2agent-unit42-2026-10",
  "name": "C2Agent LLM-driven Windows implant",
  "date_reported": "2026-10-09",
  "researcher": "Jingwen Shi, Gaurav C G, Mark Lim, Tyler Halfpop, Ashutosh Chitwadgi",
  "org": "Unit 42 (Palo Alto Networks)",
  "sources": [
    {
      "url": "https://github.com/PaloAltoNetworks/Unit42-timely-threat-intel/blob/main/2026-10-09-C2Agent-malware.txt"
    }
  ],
  "magicsword_coverage": [],
  "borderline": false,
  "borderline_note": null,
  "ai_role": "ai_infra_as_c2",
  "ai_role_secondary": null,
  "ai_role_detail": null,
  "technique_summary": "Go Windows implant where an LLM on the C2 side drives the host: operator sends a plain-language goal; implant builds a system prompt (tools, MCP, skills, memory), model returns one JSON tool call, agent executes and feeds TOOL RESULT back, loop until done (default 40 steps). Built-in MCP client defaults to Firecrawl public MCP; C2 can add MCP servers and markdown skill playbooks; skill_save lets the LLM propose new skills. From v2.3 can hand tasks to AI coding CLIs already on the host; v2.4 adds Cobalt Strike BOF loader and self-upgrade.",
  "initial_access": "unknown (two-stage: Go stager downloads PE from agent.x90.dev/data/copilot.txt)",
  "ai_tooling_and_models": [
    "LLM tool loop via C2 (proxied /llm/.../v1/chat/completions)",
    "MCP client (default mcp.firecrawl.dev)",
    "Markdown skill playbooks from C2",
    "Host AI coding CLI handoff (v2.3+)",
    "aibox.x90.dev (v0 operator LLM endpoint only)"
  ],
  "other_tooling": [
    "LOLBin templates (cred dump, shadow copies, free-space wipe)",
    "Process injection",
    "Native AD recon",
    "Exfil",
    "TCP pivot",
    "Cobalt Strike BOFs (v2.4)",
    "Startup .lnk / Run key / onlogon task"
  ],
  "victims_and_sectors": "unknown (no victim orgs named in Unit 42 timely intel note)",
  "regions": [
    "unknown"
  ],
  "attribution_or_cluster": "unknown; 32 samples ~1 month active development",
  "scale_and_metrics": "32 samples 2026-09-10 to 2026-10-08 (25 agents, 8 major versions in 29 days; 7 stager v7); payload URL live serving v2.4 as of 2026-10-09",
  "prevention_angle": "LOLSkills surface for skill/MCP abuse; hunt AI coding CLI handoff",
  "related_magicsword_projects": [
    "LOLSkills"
  ],
  "source_notes": null,
  "last_reviewed": "2026-10-20",
  "attack_flow": {
    "nodes": [
      {
        "id": "step-1",
        "title": "Stager downloads the implant",
        "description": "The reported two-stage chain fetches a Windows PE; the original delivery route is unknown.",
        "kind": "access",
        "source_urls": [
          "https://github.com/PaloAltoNetworks/Unit42-timely-threat-intel/blob/main/2026-10-09-C2Agent-malware.txt"
        ]
      },
      {
        "id": "step-2",
        "title": "LLM receives an operator goal",
        "description": "A system prompt describes available tools, MCP connections, skills and memory.",
        "kind": "ai",
        "source_urls": [
          "https://github.com/PaloAltoNetworks/Unit42-timely-threat-intel/blob/main/2026-10-09-C2Agent-malware.txt"
        ]
      },
      {
        "id": "step-3",
        "title": "Implant executes a tool call",
        "description": "The model returns a JSON tool call for the implant to execute.",
        "kind": "execution",
        "source_urls": [
          "https://github.com/PaloAltoNetworks/Unit42-timely-threat-intel/blob/main/2026-10-09-C2Agent-malware.txt"
        ]
      },
      {
        "id": "step-4",
        "title": "Tool results return to the model",
        "description": "Results feed the next decision in an iterative loop; later versions can delegate to installed AI CLIs.",
        "kind": "ai",
        "source_urls": [
          "https://github.com/PaloAltoNetworks/Unit42-timely-threat-intel/blob/main/2026-10-09-C2Agent-malware.txt"
        ]
      }
    ],
    "edges": [
      {
        "source": "step-1",
        "target": "step-2"
      },
      {
        "source": "step-2",
        "target": "step-3"
      },
      {
        "source": "step-3",
        "target": "step-4"
      },
      {
        "source": "step-4",
        "target": "step-2",
        "label": "Next decision",
        "loop": true
      }
    ]
  },
  "atlas_techniques": [
    {
      "id": "AML.T0102",
      "rationale": "An LLM returns tool calls in response to intrusion goals.",
      "source_urls": [
        "https://github.com/PaloAltoNetworks/Unit42-timely-threat-intel/blob/main/2026-10-09-C2Agent-malware.txt"
      ]
    },
    {
      "id": "AML.T0103",
      "rationale": "The implant runs an AI agent with MCP and host-side tools.",
      "source_urls": [
        "https://github.com/PaloAltoNetworks/Unit42-timely-threat-intel/blob/main/2026-10-09-C2Agent-malware.txt"
      ]
    }
  ],
  "category": "adversary-use-of-ai",
  "attribution": {
    "actor_name": "unknown"
  },
  "ai_involvement_confidence": "confirmed",
  "detections": [
    {
      "name": "Review attack-specific indicators",
      "format": "guidance",
      "notes": "Monitor for the techniques described in C2Agent LLM-driven Windows implant. Refer to sources for specific IOCs and TTPs."
    }
  ]
}