Communitygithub.com

マルウェア自動提出・分析パイプラインの構築

疑わしいファイルを自動収集し、ハッシュルックアップで事前スクリーニング、サンドボックスで動的解析を行い、判定結果とIOCをSIEMに送信するパイプラインを構築するスキルです。

マルウェア自動提出・分析パイプラインの構築 とは?

このスキルは、SOCチームやセキュリティアナリストが、大量の不審ファイルアラートを効率的に処理するための自動化パイプラインを構築する方法を提供します。高負荷な手動サンドボックス提出のボトルネックを解消し、エンドポイントやメールゲートウェイから隔離されたファイルを自動収集します。Pythonと主要なAPI(VirusTotal、MalwareBazaar、Cuckoo Sandbox、Joe Sandbox)を活用し、ハッシュによる事前スクリーニングで既知の脅威や無害なファイルを除外した後、未知のファイルのみをサンドボックスで動的解析します。最終的に、Cuckooのスコアリングとウイルス検出率を組み合わせた総合判定を生成し、抽出されたIOC(IP、ドメイン、URL、ハッシュ、レジストリキーなど)をSplunkなどのSIEMやブロックリストに自動プッシュします。このパイプラインは、Python 3.8以上の環境で動作し、隔離された解析ネットワーク上での使用が前提です。

対応~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/mukul975/Anthropic-Cybersecurity-Skills/tree/main/skills/building-automated-malware-submission-pipeline

お気に入りのAIに質問する

このエージェントスキルを事前に読み込んだ状態で新しいチャットを開きます。

ドキュメント

マルウェア自動提出・分析パイプラインの構築 は何をしますか?

使用タイミング

以下の状況でこのスキルを使用します:

  • SOCチームがサンドボックス分析を必要とする大量の不審ファイルアラートに直面している場合
  • 手動サンドボックス提出がアラートトリアージのワークフローにボトルネックを生み出している場合
  • エンドポイントおよびメールセキュリティツールが自動判定を必要とするファイルを隔離している場合
  • インシデント対応において、迅速なマルウェアファミリーの特定とIOC抽出が必要な場合

使用しないでください:本番環境でライブのマルウェアサンプルを分析する場合 — 常に隔離されたサンドボックスインフラを使用してください。

前提条件

  • サンドボックス環境:Cuckoo Sandbox、Joe Sandbox、Any.Run、またはVMRay
  • VirusTotal APIキー(提出はEnterprise、ルックアップは無料で可)
  • 既知マルウェアルックアップのためのMalwareBazaar APIアクセス
  • ファイル収集メカニズム:EDR隔離API、メールゲートウェイエクスポート、ネットワークキャプチャ
  • Python 3.8以上、requests、vt-py、pefileライブラリ
  • 本番接続のない隔離された分析ネットワーク

ワークフロー

ステップ 1: ファイル収集パイプラインの構築

複数のソースから不審ファイルを収集します:

import requests
import hashlib
import os
from pathlib import Path
from datetime import datetime

class MalwareCollector:
    def __init__(self, quarantine_dir="/opt/malware_quarantine"):
        self.quarantine_dir = Path(quarantine_dir)
        self.quarantine_dir.mkdir(exist_ok=True)

    def collect_from_edr(self, edr_api_url, api_token):
        """CrowdStrike Falconから隔離ファイルを取得"""
        headers = {"Authorization": f"Bearer {api_token}"}

        # 最近の隔離イベントを取得
        response = requests.get(
            f"{edr_api_url}/quarantine/queries/quarantined-files/v1",
            headers=headers,
            params={"filter": "state:'quarantined'", "limit": 50}
        )
        file_ids = response.json()["resources"]

        for file_id in file_ids:
            # 隔離ファイルをダウンロード
            dl_response = requests.get(
                f"{edr_api_url}/quarantine/entities/quarantined-files/v1",
                headers=headers,
                params={"ids": file_id}
            )
            file_data = dl_response.content
            sha256 = hashlib.sha256(file_data).hexdigest()

            filepath = self.quarantine_dir / f"{sha256}.sample"
            filepath.write_bytes(file_data)
            yield {"sha256": sha256, "path": str(filepath), "source": "edr"}

    def collect_from_email_gateway(self, smtp_quarantine_path):
        """メールゲートウェイの隔離から添付ファイルを取得"""
        import email
        from email import policy

        for eml_file in Path(smtp_quarantine_path).glob("*.eml"):
            msg = email.message_from_binary_file(
                eml_file.open("rb"), policy=policy.default
            )
            for attachment in msg.iter_attachments():
                content = attachment.get_content()
                if isinstance(content, str):
                    content = content.encode()
                sha256 = hashlib.sha256(content).hexdigest()
                filename = attachment.get_filename() or "unknown"

                filepath = self.quarantine_dir / f"{sha256}.sample"
                filepath.write_bytes(content)
                yield {
                    "sha256": sha256,
                    "path": str(filepath),
                    "source": "email",
                    "original_filename": filename,
                    "sender": msg["From"],
                    "subject": msg["Subject"]
                }

    def compute_hashes(self, filepath):
        """ファイルのMD5、SHA1、SHA256を計算"""
        with open(filepath, "rb") as f:
            content = f.read()
        return {
            "md5": hashlib.md5(content).hexdigest(),
            "sha1": hashlib.sha1(content).hexdigest(),
            "sha256": hashlib.sha256(content).hexdigest(),
            "size": len(content)
        }

ステップ 2: ハッシュルックアップによる事前スクリーニング

サンドボックスに提出する前に、ファイルが既知かどうかを確認します:

import vt

class MalwarePreScreener:
    def __init__(self, vt_api_key, mb_api_url="https://mb-api.abuse.ch/api/v1/"):
        self.vt_client = vt.Client(vt_api_key)
        self.mb_api_url = mb_api_url

    def check_virustotal(self, sha256):
        """VirusTotalでハッシュをルックアップ"""
        try:
            file_obj = self.vt_client.get_object(f"/files/{sha256}")
            stats = file_obj.last_analysis_stats
            return {
                "found": True,
                "malicious": stats.get("malicious", 0),
                "suspicious": stats.get("suspicious", 0),
                "undetected": stats.get("undetected", 0),
                "total": sum(stats.values()),
                "threat_label": getattr(file_obj, "popular_threat_classification", {}).get(
                    "suggested_threat_label", "Unknown"
                ),
                "type": getattr(file_obj, "type_description", "Unknown")
            }
        except vt.APIError:
            return {"found": False}

    def check_malwarebazaar(self, sha256):
        """MalwareBazaarでハッシュをルックアップ"""
        response = requests.post(
            self.mb_api_url,
            data={"query": "get_info", "hash": sha256}
        )
        data = response.json()
        if data["query_status"] == "ok":
            entry = data["data"][0]
            return {
                "found": True,
                "signature": entry.get("signature", "Unknown"),
                "tags": entry.get("tags", []),
                "file_type": entry.get("file_type", "Unknown"),
                "first_seen": entry.get("first_seen", "Unknown")
            }
        return {"found": False}

    def pre_screen(self, sha256):
        """すべての事前スクリーニングチェックを実行"""
        vt_result = self.check_virustotal(sha256)
        mb_result = self.check_malwarebazaar(sha256)

        verdict = "UNKNOWN"
        if vt_result["found"] and vt_result.get("malicious", 0) > 10:
            verdict = "KNOWN_MALICIOUS"
        elif vt_result["found"] and vt_result.get("malicious", 0) == 0:
            verdict = "LIKELY_CLEAN"

        return {
            "sha256": sha256,
            "virustotal": vt_result,
            "malwarebazaar": mb_result,
            "pre_screen_verdict": verdict,
            "needs_sandbox": verdict == "UNKNOWN"
        }

    def close(self):
        self.vt_client.close()

ステップ 3: 動的分析のためにサンドボックスに提出

Cuckoo Sandboxへの提出:

class SandboxSubmitter:
    def __init__(self, cuckoo_url="http://cuckoo.internal:8090"):
        self.cuckoo_url = cuckoo_url

    def submit_to_cuckoo(self, filepath, timeout=300):
        """ファイルをCuckoo Sandboxに提出"""
        with open(filepath, "rb") as f:
            response = requests.post(
                f"{self.cuckoo_url}/tasks/create/file",
                files={"file": f},
                data={
                    "timeout": timeout,
                    "options": "procmemdump=yes,route=none",
                    "priority": 2,
                    "machine": "win10_x64"
                }
            )
        task_id = response.json()["task_id"]
        return task_id

    def wait_for_analysis(self, task_id, poll_interval=30, max_wait=600):
        """サンドボックス分析の完了を待機"""
        import time
        elapsed = 0
        while elapsed < max_wait:
            response = requests.get(f"{self.cuckoo_url}/tasks/view/{task_id}")
            status = response.json()["task"]["status"]
            if status == "reported":
                return self.get_report(task_id)
            elif status == "failed_analysis":
                return {"error": "Analysis failed"}
            time.sleep(poll_interval)
            elapsed += poll_interval
        return {"error": "Analysis timed out"}

    def get_report(self, task_id):
        """分析レポートを取得"""
        response = requests.get(f"{self.cuckoo_url}/tasks/report/{task_id}")
        report = response.json()

        # 主要な指標を抽出
        return {
            "task_id": task_id,
            "score": report.get("info", {}).get("score", 0),
            "signatures": [
                {"name": s["name"], "severity": s["severity"], "description": s["description"]}
                for s in report.get("signatures", [])
            ],
            "network": {
                "dns": [d["request"] for d in report.get("network", {}).get("dns", [])],
                "http": [
                    {"url": h["uri"], "method": h["method"]}
                    for h in report.get("network", {}).get("http", [])
                ],
                "hosts": report.get("network", {}).get("hosts", [])
            },
            "dropped_files": [
                {"name": f["name"], "sha256": f["sha256"], "size": f["size"]}
                for f in report.get("dropped", [])
            ],
            "processes": [
                {"name": p["process_name"], "pid": p["pid"], "command_line": p.get("command_line", "")}
                for p in report.get("behavior", {}).get("processes", [])
            ],
            "registry_keys": [
                k for k in report.get("behavior", {}).get("summary", {}).get("regkey_written", [])
            ]
        }

    def submit_to_joesandbox(self, filepath, joe_api_key, joe_url="https://jbxcloud.joesecurity.org/api"):
        """Joe Sandbox Cloudに提出"""
        with open(filepath, "rb") as f:
            response = requests.post(
                f"{joe_url}/v2/submission/new",
                headers={"Authorization": f"Bearer {joe_api_key}"},
                files={"sample": f},
                data={
                    "systems": "w10_64",
                    "internet-access": False,
                    "report-cache": True
                }
            )
        return response.json()["data"]["webid"]

ステップ 4: IOCの抽出と判定の生成

class VerdictGenerator:
    def __init__(self):
        self.malicious_threshold = 7  # Cuckooのスコア閾値

    def generate_verdict(self, pre_screen, sandbox_report):
        """事前スクリーニングとサンドボックス結果を組み合わせて最終判定を生成"""
        iocs = {
            "ips": [],
            "domains": [],
            "urls": [],
            "hashes": [],
            "registry_keys": [],
            "files_dropped": []
        }

        # サンドボックスレポートからIOCを抽出
        if sandbox_report:
            iocs["domains"] = sandbox_report.get("network", {}).get("dns", [])
            iocs["ips"] = sandbox_report.get("network", {}).get("hosts", [])
            iocs["urls"] = [
                h["url"] for h in sandbox_report.get("network", {}).get("http", [])
            ]
            iocs["hashes"] = [
                f["sha256"] for f in sandbox_report.get("dropped_files", [])
            ]
            iocs["registry_keys"] = sandbox_report.get("registry_keys", [])[:10]
            iocs["files_dropped"] = sandbox_report.get("dropped_files", [])

        # 判定を決定
        vt_malicious = pre_screen.get("virustotal", {}).get("malicious", 0)
        sandbox_score = sandbox_report.get("score", 0) if sandbox_report else 0
        sig_count = len(sandbox_report.get("signatures", [])) if sandbox_report else 0

        combined_score = (vt_malicious * 2) + (sandbox_score * 10) + (sig_count * 5)

        if combined_score >= 100:
            verdict = "MALICIOUS"
            confidence = "HIGH"
        elif combined_score >= 50:
            verdict = "SUSPICIOUS"
            confidence = "MEDIUM"
        elif combined_score >= 20:
            verdict = "POTENTIALLY_UNWANTED"
            confidence = "LOW"
        else:
            verdict = "CLEAN"
            confidence = "HIGH"

        return {
            "verdict": verdict,
            "confidence": confidence,
            "combined_score": combined_score,
            "iocs": iocs,
            "vt_detections": vt_malicious,
            "sandbox_score": sandbox_score,
            "signatures": sandbox_report.get("signatures", []) if sandbox_report else []
        }

ステップ 5: 結果をSIEMにプッシュ

def push_to_splunk(verdict_result, splunk_url, splunk_token):
    """マルウェア分析の判定をSplunk HECに送信"""
    import json

    event = {
        "sourcetype": "malware_analysis",
        "source": "malware_pipeline",
        "event": {
            "sha256": verdict_result["sha256"],
            "verdict": verdict_result["verdict"],
            "confidence": verdict_result["confidence"],
            "score": verdict_result["combined_score"],
            "vt_detections": verdict_result["vt_detections"],
            "sandbox_score": verdict_result["sandbox_score"],
            "malware_family": verdict_result.get("threat_label", "Unknown"),
            "iocs": verdict_result["iocs"],
            "signatures": [s["name"] for s in verdict_result["signatures"]]
        }
    }

    response = requests.post(
        f"{splunk_url}/services/collector/event",
        headers={
            "Authorization": f"Splunk {splunk_token}",
            "Content-Type": "application/json"
        },
        json=event,
        verify=not os.environ.get("SKIP_TLS_VERIFY", "").lower() == "true",  # ラボ環境での自己署名証明書用にSKIP_TLS_VERIFY=trueを設定
    )
    return response.status_code == 200

def push_iocs_to_blocklist(iocs, firewall_api):
    """抽出したIOCをブロックインフラにプッシュ"""
    for ip in iocs.get("ips", []):
        requests.post(
            f"{firewall_api}/block",
            json={"type": "ip", "value": ip, "action": "block", "source": "malware_pipeline"}
        )
    for domain in iocs.get("domains", []):
        requests.post(
            f"{firewall_api}/block",
            json={"type": "domain", "value": domain, "action": "sinkhole", "source": "malware_pipeline"}
        )

ステップ 6: パイプライン全体をオーケストレーション

def run_malware_pipeline():
    """完全なマルウェア分析パイプラインを実行"""
    # 設定
    VT_API_KEY = os.environ["VT_API_KEY"]
    SPLUNK_URL = os.environ.get("SPLUNK_URL")

Individual skills in this repo

This repo contains 20 individual skills — each has its own dedicated page.

abusing-dpapi-for-credential-access

Extract and decrypt Windows DPAPI-protected secrets (Credential Manager, browser logins/cookies, Wi-Fi credentials, KeePass keys) online or offline using SharpDPAPI, SharpChrome, Mimikatz, or Impacket

abusing-shadow-credentials-for-privesc

Take over Active Directory accounts by writing attacker-controlled public keys to msDS-KeyCredentialLink (Shadow Credentials) with pyWhisker, Whisker, or Certipy, then authenticate via PKINIT to recover the target

achieving-cmmc-level-2-compliance

>-

acquiring-disk-image-with-dd-and-dcfldd

Create forensically sound bit-for-bit disk images with dd or dcfldd on a Linux forensic workstation, preserving evidence integrity through hash verification (MD5/SHA) during acquisition. Use when imaging a suspect drive, USB device, or memory card for investigation, preserving volatile disk evidence during incident response, or producing a verified copy for legal or law-enforcement proceedings before any destructive analysis.

analyzing-active-directory-acl-abuse

Detect dangerous ACL misconfigurations in Active Directory using ldap3

analyzing-android-malware-with-apktool

Perform static analysis of Android APK malware using apktool for resource decompilation, jadx for Java source recovery, and androguard for manifest inspection, dangerous permission-combination detection, and identification of obfuscated code, dynamic code loading, and reflection-based API calls. Use to statically triage a suspicious APK without executing it or to build mobile malware detection rules.

analyzing-api-gateway-access-logs

Parses API Gateway access logs (AWS API Gateway, Kong, Nginx) to detect

analyzing-apt-group-with-mitre-navigator

Query ATT&CK data with attackcti, mitreattack-python, and stix2, then build MITRE ATT&CK Navigator layers and multi-layer heatmap overlays mapping one or more APT groups

analyzing-azure-activity-logs-for-threats

Queries Azure Monitor activity logs and sign-in logs via azure-monitor-query

analyzing-bootkit-and-rootkit-samples

Analyzes bootkit and advanced rootkit malware infecting the Master

analyzing-browser-forensics-with-hindsight

Parse Chromium-based browser databases with Hindsight to extract and correlate browsing history, downloads, cookies, cached content, autofill data, saved passwords, and extensions from Chrome, Edge, Brave, Opera, and Vivaldi into a unified timeline (XLSX, JSON, or SQLite output). Use during incident response, insider-threat investigations, or criminal cases when you need to reconstruct a user

analyzing-campaign-attribution-evidence

Systematically evaluate cyber-campaign evidence to attribute an operation to a threat actor, using the Diamond Model and Analysis of Competing Hypotheses (ACH) to weigh infrastructure overlaps, TTP consistency, malware code similarity, and timing/language artifacts into confidence-weighted attribution assessments. Use when an incident investigation needs a defensible attribution confidence level.

analyzing-certificate-transparency-for-phishing

Monitor Certificate Transparency logs using crt.sh and Certstream to

analyzing-cloud-storage-access-patterns

Detect abnormal access in AWS S3, GCS, and Azure Blob Storage by analyzing CloudTrail Data Events, GCS audit logs, and Azure Storage Analytics for after-hours bulk downloads, new-IP access, and API-call spikes (e.g. GetObject) via statistical baselines and time-series anomaly detection. Use when investigating suspected cloud data exfiltration or building related detection rules.

analyzing-cobalt-strike-beacon-configuration

Extract and analyze Cobalt Strike beacon configuration from PE files

analyzing-cobaltstrike-malleable-c2-profiles

Parse and analyze Cobalt Strike Malleable C2 profiles with dissect.cobaltstrike (profiles and beacon-payload configs) and pyMalleableC2 (AST parsing) to extract HTTP/DNS transforms, URIs, headers, sleep/jitter, and injection behavior, then generate network detection signatures. Use when reverse-engineering a captured malleable profile or building detections against Cobalt Strike Beacon traffic.

analyzing-command-and-control-communication

Analyzes malware C2 communication over HTTP, HTTPS, DNS, and custom

analyzing-cyber-kill-chain

Analyzes intrusion activity against the Lockheed Martin Cyber Kill Chain

analyzing-disk-image-with-autopsy

Perform comprehensive forensic analysis of raw (dd), E01, or AFF disk images with Autopsy and The Sleuth Kit, recovering deleted files, examining metadata and embedded artifacts, keyword searching, and building investigation timelines with visual reports. Use for structured analysis of a forensic disk image or when stakeholders need visual reports from evidence.

analyzing-dns-logs-for-exfiltration

Analyzes DNS query logs to detect data exfiltration via DNS tunneling,

関連スキル