本文へ移動
cccskills
無料GitHub で公開

building-ioc-enrichment-pipeline-with-opencti

OpenCTI is an open-source platform for managing cyber threat intelligence knowledge, built on STIX 2.1 as its native data model. This skill covers building an automated IOC enrichment pipeline using O

インストール方法を見る

含まれるファイル(8)

  • SKILL.md9.3 KB
  • assets/template.md2.0 KB
  • LICENSE11.0 KB
  • references/api-reference.md1.8 KB
  • references/standards.md3.1 KB
  • references/workflows.md5.0 KB
  • scripts/agent.py5.7 KB
  • scripts/process.py17.3 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Building IOC Enrichment Pipeline with OpenCTI

Overview

OpenCTI is an open-source platform for managing cyber threat intelligence knowledge, built on STIX 2.1 as its native data model. This skill covers building an automated IOC enrichment pipeline using OpenCTI's connector ecosystem to enrich indicators with context from VirusTotal, Shodan, AbuseIPDB, GreyNoise, and other sources. The pipeline automatically enriches newly ingested indicators, correlates them with known threat actors and campaigns, and scores them for analyst prioritization.

When to Use

  • When deploying or configuring building ioc enrichment pipeline with opencti capabilities in your environment
  • When establishing security controls aligned to compliance requirements
  • When building or improving security architecture for this domain
  • When conducting security assessments that require this implementation

Prerequisites

  • Docker and Docker Compose for OpenCTI deployment
  • Python 3.9+ with pycti library
  • API keys for enrichment services: VirusTotal, Shodan, AbuseIPDB, GreyNoise
  • Understanding of STIX 2.1 data model and relationships
  • ElasticSearch or OpenSearch for OpenCTI backend
  • RabbitMQ or Redis for connector messaging

Key Concepts

OpenCTI Architecture

OpenCTI uses a GraphQL API frontend backed by ElasticSearch for storage and Redis/RabbitMQ for connector communication. Data is natively stored as STIX 2.1 objects with relationships. Connectors are categorized as: External Import (feed ingestion), Internal Import (file parsing), Internal Enrichment (context addition), and Stream (real-time export).

Enrichment Connector Model

Internal enrichment connectors are triggered automatically when new observables are created or manually by analysts. Each connector receives STIX objects, queries external services, and returns STIX 2.1 bundles that augment the original observable with additional context, labels, and relationships.

Confidence Scoring

OpenCTI uses a 0-100 confidence scale for indicators. Enrichment connectors can update confidence scores based on external validation: VirusTotal detection ratios, Shodan exposure data, AbuseIPDB report counts, and GreyNoise classification results.

Workflow

Step 1: Deploy OpenCTI with Docker Compose

# docker-compose.yml (key services)
version: '3'
services:
  opencti:
    image: opencti/platform:6.4.4
    environment:
      - APP__PORT=8080
      - APP__ADMIN__EMAIL=admin@opencti.io
      - APP__ADMIN__PASSWORD=ChangeMeNow
      - APP__ADMIN__TOKEN=your-admin-token-uuid
      - ELASTICSEARCH__URL=http://elasticsearch:9200
      - MINIO__ENDPOINT=minio
      - RABBITMQ__HOSTNAME=rabbitmq
    ports:
      - "8080:8080"
    depends_on:
      - elasticsearch
      - minio
      - rabbitmq
      - redis

  connector-virustotal:
    image: opencti/connector-virustotal:6.4.4
    environment:
      - OPENCTI_URL=http://opencti:8080
      - OPENCTI_TOKEN=your-admin-token-uuid
      - CONNECTOR_ID=connector-virustotal-id
      - CONNECTOR_NAME=VirusTotal
      - CONNECTOR_SCOPE=StixFile,Artifact,IPv4-Addr,Domain-Name,Url
      - CONNECTOR_AUTO=true
      - VIRUSTOTAL_TOKEN=your-vt-api-key
      - VIRUSTOTAL_MAX_TLP=TLP:AMBER

  connector-shodan:
    image: opencti/connector-shodan:6.4.4
    environment:
      - OPENCTI_URL=http://opencti:8080
      - OPENCTI_TOKEN=your-admin-token-uuid
      - CONNECTOR_ID=connector-shodan-id
      - CONNECTOR_NAME=Shodan
      - CONNECTOR_SCOPE=IPv4-Addr
      - CONNECTOR_AUTO=true
      - SHODAN_TOKEN=your-shodan-api-key
      - SHODAN_MAX_TLP=TLP:AMBER

  connector-abuseipdb:
    image: opencti/connector-abuseipdb:6.4.4
    environment:
      - OPENCTI_URL=http://opencti:8080
      - OPENCTI_TOKEN=your-admin-token-uuid
      - CONNECTOR_ID=connector-abuseipdb-id
      - CONNECTOR_NAME=AbuseIPDB
      - CONNECTOR_SCOPE=IPv4-Addr
      - CONNECTOR_AUTO=true
      - ABUSEIPDB_API_KEY=your-abuseipdb-key

Step 2: Build Custom Enrichment Connector

import os
from pycti import OpenCTIConnectorHelper, get_config_variable
from stix2 import (
    Bundle, Indicator, Note, Relationship,
    IPv4Address, DomainName
)
import requests


class CustomEnrichmentConnector:
    def __init__(self):
        config = {
            "opencti": {
                "url": os.environ.get("OPENCTI_URL"),
                "token": os.environ.get("OPENCTI_TOKEN"),
            },
            "connector": {
                "id": os.environ.get("CONNECTOR_ID"),
                "name": "CustomEnrichment",
                "scope": "IPv4-Addr,Domain-Name,Url",
                "auto": True,
                "type": "INTERNAL_ENRICHMENT",
            },
        }
        self.helper = OpenCTIConnectorHelper(config)
        self.helper.listen(self._process_message)

    def _process_message(self, data):
        entity_id = data["entity_id"]
        stix_object = self.helper.api.stix_cyber_observable.read(id=entity_id)

        if not stix_object:
            return "Observable not found"

        observable_type = stix_object["entity_type"]
        observable_value = stix_object.get("value", "")

        enrichment_results = []

        if observable_type == "IPv4-Addr":
            enrichment_results = self._enrich_ip(observable_value, entity_id)
        elif observable_type == "Domain-Name":
            enrichment_results = self._enrich_domain(observable_value, entity_id)

        if enrichment_results:
            bundle = Bundle(objects=enrichment_results, allow_custom=True)
            self.helper.send_stix2_bundle(bundle.serialize())

        return "Enrichment completed"

    def _enrich_ip(self, ip_address, entity_id):
        """Enrich IP address with GreyNoise, AbuseIPDB context."""
        objects = []

        # GreyNoise Community API
        try:
            gn_response = requests.get(
                f"https://api.greynoise.io/v3/community/{ip_address}",
                headers={"key": os.environ.get("GREYNOISE_API_KEY")},
                timeout=30,
            )
            if gn_response.status_code == 200:
                gn_data = gn_response.json()
                classification = gn_data.get("classification", "unknown")
                noise = gn_data.get("noise", False)
                riot = gn_data.get("riot", False)

                note_content = (
                    f"## GreyNoise Enrichment\n"
                    f"- Classification: {classification}\n"
                    f"- Internet Noise: {noise}\n"
                    f"- RIOT (Benign Service): {riot}\n"
                    f"- Name: {gn_data.get('name', 'N/A')}\n"
                    f"- Last Seen: {gn_data.get('last_seen', 'N/A')}"
                )

                note = Note(
                    content=note_content,
                    object_refs=[entity_id],
                    abstract=f"GreyNoise: {classification}",
                    allow_custom=True,
                )
                objects.append(note)

                # Add labels based on classification
                if classification == "malicious":
                    self.helper.api.stix_cyber_observable.add_label(
                        id=entity_id, label_name="greynoise:malicious"
                    )
                elif riot:
                    self.helper.api.stix_cyber_observable.add_label(
                        id=entity_id, label_name="greynoise:benign-service"
                    )

        except Exception as e:
            self.helper.log_error(f"GreyNoise enrichment failed: {e}")

        return objects

    def _enrich_domain(self, domain, entity_id):
        """Enrich domain with WHOIS and DNS context."""
        objects = []

        try:
            # Use SecurityTrails API for domain enrichment
            st_response = requests.get(
                f"https://api.securitytrails.com/v1/domain/{domain}",
                headers={"APIKEY": os.environ.get("SECURITYTRAILS_API_KEY")},
                timeout=30,
            )
            if st_response.status_code == 200:
                st_data = st_response.json()
                current_dns = st_data.get("current_dns", {})

                a_records = [
                    r.get("ip") for r in current_dns.get("a", {}).get("values", [])
                ]

                note_content = (
                    f"## SecurityTrails Enrichment\n"
                    f"- A Records: {', '.join(a_records)}\n"
                    f"- Alexa Rank: {st_data.get('alexa_rank', 'N/A')}\n"
                    f"- Hostname: {st_data.get('hostname', 'N/A')}"
                )

                note = Note(
                    content=note_content,
                    object_refs=[entity_id],
                    abstract=f"SecurityTrails: {domain}",
                    allow_custom=True,
                )
                objects.append(note)

        except Exception as e:
            self.helper.log_error(f"SecurityTrails enrichment failed: {e}")

        return objects


if __name__ == "__main__":
    connector = CustomEnrichmentConnector()

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Create forensically sound bit-for-bit disk images using dd and dcfldd while preserving evidence integrity through hash verification.

日本語の概要は準備中です。原文の説明を表示しています。

micsapp/micstec-skills42026年3月20日 更新

Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.

日本語の概要は準備中です。原文の説明を表示しています。

micsapp/micstec-skills42026年3月20日 更新

Detect dangerous ACL misconfigurations in Active Directory using ldap3 to identify GenericAll, WriteDACL, and WriteOwner abuse paths

日本語の概要は準備中です。原文の説明を表示しています。

micsapp/micstec-skills42026年3月20日 更新

Perform static analysis of Android APK malware samples using apktool for decompilation, jadx for Java source recovery, and androguard for permission analysis, manifest inspection, and suspicious API call detection.

日本語の概要は準備中です。原文の説明を表示しています。

micsapp/micstec-skills42026年3月20日 更新

Parses API Gateway access logs (AWS API Gateway, Kong, Nginx) to detect BOLA/IDOR attacks, rate limit bypass, credential scanning, and injection attempts. Uses pandas for statistical analysis of request patterns and anomaly detection. Use when investigating API abuse or building API-specific threat detection rules.

日本語の概要は準備中です。原文の説明を表示しています。

micsapp/micstec-skills42026年3月20日 更新

Analyze advanced persistent threat (APT) group techniques using MITRE ATT&CK Navigator to create layered heatmaps of adversary TTPs for detection gap analysis and threat-informed defense.

日本語の概要は準備中です。原文の説明を表示しています。

micsapp/micstec-skills42026年3月20日 更新

micsapp のスキルをすべて見る

このスキルの問題を報告する