Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build an automated pipeline to defang indicators of compromise (URLs, IPs, domains, emails) for safe sharing and distribute them in STIX format through TAXII feeds and threat intelligence platforms.
.claude/skills/building-ioc-defanging-and-sharing-pipeline/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flashlowest | 91% | 11 |
| gemini-3.1-pro-preview | 100% | 2 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | — | — |
| case-17 | ✗→✓ | ▲ Improved | — | — |
| case-05 | ✗→✓ | ▲ Improved | — | — |
| case-07 | ✗→✓ | ▲ Improved | — | — |
| case-08 | ✓→✓ | = Same ✓ | — | — |
IOC defanging modifies potentially malicious indicators (URLs, IP addresses, domains, email addresses) to prevent accidental clicks or execution while preserving readability for analysis and sharing. This skill covers building an automated pipeline that ingests raw IOCs from multiple sources, normalizes and deduplicates them, applies defanging for safe human consumption, converts them to STIX 2.1 format for machine consumption, and distributes through TAXII servers, MISP instances, and email reports.
defang, ioc-fanger, stix2, requests, validators librariesDefanging replaces active protocol and domain components to prevent execution: http:// becomes hxxp://, https:// becomes hxxps://, dots in domains/IPs become [.], @ in emails becomes [@]. This is critical for sharing IOCs in reports, emails, Slack channels, and paste sites where auto-linking could trigger network connections to malicious infrastructure.
Raw IOCs from different sources come in inconsistent formats. Normalization involves converting to lowercase, removing trailing slashes and whitespace, extracting domains from URLs, resolving URL encoding, validating format correctness, and deduplicating across sources.
STIX patterns express IOCs in a standardized format: [ipv4-addr:value = '203.0.113.1'], [domain-name:value = 'malicious.example.com'], [url:value = 'http://evil.com/payload'], [file:hashes.'SHA-256' = 'abc123...']. Each indicator includes valid_from, indicator_types, confidence, and optional TLP markings.
pythonimport re import hashlib from urllib.parse import urlparse, unquote from datetime import datetime class IOCExtractor: """Extract and normalize IOCs from text.""" PATTERNS = { "ipv4": r'\b(?:(?:25[0-5]|2[0-4]\d|1\d{2}|[1-9]?\d)\.){3}(?:25[0-5]|2[0-4]\d|1\d{2}|[1-9]?\d)\b', "domain": r'\b(?:[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?\.)+[a-zA-Z]{2,}\b', "url": r'https?://[^\s<>"{}|\\^`\[\]]+', "email": r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b', "md5": r'\b[a-fA-F0-9]{32}\b', "sha1": r'\b[a-fA-F0-9]{40}\b', "sha256": r'\b[a-fA-F0-9]{64}\b', } WHITELIST_DOMAINS = { "google.com", "microsoft.com", "amazon.com", "github.com", "cloudflare.com", "akamai.com", "example.com", } def extract_from_text(self, text): """Extract all IOC types from free text.""" # Refang any already-defanged indicators first text = self._refang(text) iocs = {"ipv4": set(), "domain": set(), "url": set(), "email": set(), "md5": set(), "sha1": set(), "sha256": set()} for ioc_type, pattern in self.PATTERNS.items(): matches = re.findall(pattern, text) for match in matches: normalized = self._normalize(match, ioc_type) if normalized and not self._is_whitelisted(normalized, ioc_type): iocs[ioc_type].add(normalized) # Remove domains that are part of URLs url_domains = set() for url in iocs["url"]: parsed = urlparse(url) url_domains.add(parsed.netloc) iocs["domain"] -= url_domains total = sum(len(v) for v in iocs.values()) print(f"[+] Extracted {total} unique IOCs from text") return {k: sorted(v) for k, v in iocs.items()} def _refang(self, text): """Convert defanged indicators back to active form.""" text = text.replace("hxxp://", "http://").replace("hxxps://", "https://") text = text.replace("[.]", ".").replace("[@]", "@") text = text.replace("[://]", "://").replace("(.)", ".") return text def _normalize(self, value, ioc_type): """Normalize an IOC value.""" value = value.strip().lower() if ioc_type == "url": value = unquote(value).rstrip("/") elif ioc_type == "domain": value = value.rstrip(".") return value def _is_whitelisted(self, value, ioc_type): """Check if IOC is in whitelist.""" if ioc_type == "domain": return value in self.WHITELIST_DOMAINS if ioc_type == "url": parsed = urlparse(value) return parsed.netloc in self.WHITELIST_DOMAINS return False extractor = IOCExtractor() sample_text = """ Malware C2: hxxps://evil-domain[.]com/beacon Drops payload from 192.168.1.100 and contacts 10[.]0[.]0[.]1 SHA256: 275a021bbfb6489e54d471899f7db9d1663fc695ec2fe2a2c4538aabf651fd0f Phishing email from attacker[@]phishing-domain[.]com """ iocs = extractor.extract_from_text(sample_text)
pythonclass IOCDefanger: """Defang IOCs for safe sharing in reports and communications.""" def defang_url(self, url): return url.replace("http://", "hxxp://").replace("https://", "hxxps://").replace(".", "[.]") def defang_domain(self, domain): return domain.replace(".", "[.]") def defang_ip(self, ip): return ip.replace(".", "[.]") def defang_email(self, email): return email.replace("@", "[@]").replace(".", "[.]") def defang_all(self, iocs): """Defang all IOCs in a dictionary.""" defanged = {} for ioc_type, values in iocs.items(): if ioc_type == "url": defanged[ioc_type] = [self.defang_url(v) for v in values] elif ioc_type == "domain": defanged[ioc_type] = [self.defang_domain(v) for v in values] elif ioc_type == "ipv4": defanged[ioc_type] = [self.defang_ip(v) for v in values] elif ioc_type == "email": defanged[ioc_type] = [self.defang_email(v) for v in values] else: defanged[ioc_type] = values # Hashes don't need defanging return defanged def generate_sharing_report(self, iocs, defanged, report_name="IOC Report"): """Generate a human-readable defanged IOC report.""" report = f"# {report_name}\n" report += f"Generated: {datetime.now().isoformat()}\n\n" for ioc_type in ["url", "domain", "ipv4", "email", "sha256", "sha1", "md5"]: values = defanged.get(ioc_type, []) if values: report += f"## {ioc_type.upper()} ({len(values)})\n" for v in values: report += f"- `{v}`\n" report += "\n" return report defanger = IOCDefanger() defanged = defanger.defang_all(iocs) report = defanger.generate_sharing_report(iocs, defanged, "Malware Campaign IOCs") print(report)
pythonfrom stix2 import Indicator, Bundle, TLP_WHITE, TLP_GREEN, TLP_AMBER from datetime import datetime class STIXConverter: """Convert raw IOCs to STIX 2.1 Indicator objects.""" TLP_MAP = {"white": TLP_WHITE, "green": TLP_GREEN, "amber": TLP_AMBER} def iocs_to_stix(self, iocs, tlp="green", confidence=75): """Convert IOC dictionary to STIX 2.1 bundle.""" stix_objects = [] marking = self.TLP_MAP.get(tlp, TLP_GREEN) for ip in iocs.get("ipv4", []): stix_objects.append(Indicator( name=f"Malicious IP: {ip}", pattern=f"[ipv4-addr:value = '{ip}']", pattern_type="stix", valid_from=datetime.now(), indicator_types=["malicious-activity"], confidence=confidence, object_marking_refs=[marking], )) for domain in iocs.get("domain", []): stix_objects.append(Indicator( name=f"Malicious Domain: {domain}", pattern=f"[domain-name:value = '{domain}']", pattern_type="stix", valid_from=datetime.now(), indicator_types=["malicious-activity"], confidence=confidence, object_marking_refs=[marking], )) for url in iocs.get("url", []): escaped = url.replace("'", "\\'") stix_objects.append(Indicator( name=f"Malicious URL: {url[:60]}", pattern=f"[url:value = '{escaped}']", pattern_type="stix", valid_from=datetime.now(), indicator_types=["malicious-activity"], confidence=confidence, object_marking_refs=[marking], )) for sha256 in iocs.get("sha256", []): stix_objects.append(Indicator( name=f"Malicious File Hash: {sha256[:16]}...", pattern=f"[file:hashes.'SHA-256' = '{sha256}']", pattern_type="stix", valid_from=datetime.now(), indicator_types=["malicious-activity"], confidence=confidence, object_marking_refs=[marking], )) bundle = Bundle(objects=stix_objects) print(f"[+] Created STIX bundle with {len(stix_objects)} indicators") return bundle converter = STIXConverter() stix_bundle = converter.iocs_to_stix(iocs, tlp="amber", confidence=80) with open("iocs_stix_bundle.json", "w") as f: f.write(stix_bundle.serialize(pretty=True))
pythonimport requests import json class IOCDistributor: """Distribute IOCs through various channels.""" def push_to_misp(self, iocs, misp_url, misp_key, event_info): """Push IOCs to MISP as a new event.""" headers = { "Authorization": misp_key, "Content-Type": "application/json", "Accept": "application/json", } event = { "Event": { "info": event_info, "distribution": "1", # This community only "threat_level_id": "2", # Medium "analysis": "2", # Completed "Attribute": [], } } type_mapping = { "ipv4": "ip-dst", "domain": "domain", "url": "url", "email": "email-src", "md5": "md5", "sha1": "sha1", "sha256": "sha256", } for ioc_type, values in iocs.items(): misp_type = type_mapping.get(ioc_type) if misp_type: for value in values: event["Event"]["Attribute"].append({ "type": misp_type, "value": value, "category": "Network activity" if ioc_type in ("ipv4", "domain", "url") else "Payload delivery", "to_ids": True, }) resp = requests.post( f"{misp_url}/events", headers=headers, json=event, verify=not os.environ.get("SKIP_TLS_VERIFY", "").lower() == "true", # Set SKIP_TLS_VERIFY=true for self-signed certs in lab environments ) if resp.status_code == 200: event_id = resp.json().get("Event", {}).get("id", "") print(f"[+] MISP event created: {event_id}") return event_id else: print(f"[-] MISP error: {resp.status_code} - {resp.text[:200]}") return None def push_to_taxii(self, stix_bundle, taxii_url, collection_id, username, password): """Push STIX bundle to TAXII 2.1 collection.""" from taxii2client.v21 import Collection collection = Collection( f"{taxii_url}/collections/{collection_id}/", user=username, password=password, ) response = collection.add_objects(stix_bundle.serialize()) print(f"[+] TAXII: Published bundle, status: {response.status}") return response distributor = IOCDistributor() distributor.push_to_misp( iocs, misp_url="https://misp.organization.com", misp_key="YOUR_MISP_API_KEY", event_info="Malware Campaign IOCs - 2025", )
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-24 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 24 comparable cases.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.