Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
mukul975 avatar

Analyzing Indicators Of Compromise

  • 375 installs
  • 27.3k repo stars
  • Updated August 2, 2026
  • mukul975/anthropic-cybersecurity-skills

Run a structured investigation when logs or alerts surface suspicious hashes, IPs, domains, or URLs.

About

Analyzing Indicators Of Compromise is a Security agent skill for solo and indie builders who need repeatable triage when something looks wrong in production or pre-launch monitoring. Instead of improvising in chat, you hand your coding agent concrete artifacts—hashes, IPs, domains, URLs, or pasted intel—and follow a procedural path to validate severity, link related indicators, and decide what to block, rotate, or patch. It fits the Ship security shelf and Operate incident moments when you are wearing every hat and cannot afford a vague “looks fine” answer. The skill emphasizes documentation-friendly conclusions you can attach to a ticket, share with a cofounder, or feed into a deeper forensics step. It does not replace commercial EDR or paid threat feeds; it gives you agent-native procedure so investigation stays consistent under time pressure.

  • Structured workflow for analyzing file hashes, IP addresses, domains, and URLs from alerts or exports
  • Correlation guidance to tie multiple IOC types to the same incident narrative
  • Triage-oriented outputs suited to solo builders without a full SOC stack
  • Framed for agent-assisted investigation alongside your existing logs and tooling
  • Apache 2.0 skill package from the anthropic-cybersecurity-skills collection

Analyzing Indicators Of Compromise by the numbers

  • 375 all-time installs (skills.sh)
  • +20 installs in the week ending Aug 4, 2026 (Skillselion tracking)
  • Ranked #570 of 2,203 Security skills by installs in the Skillselion catalog
  • Security screen: CRITICAL risk (skills.sh audit)
  • Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/mukul975/anthropic-cybersecurity-skills --skill analyzing-indicators-of-compromise

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs375
repo stars27.3k
Security audit1 / 3 scanners passed
Last updatedAugust 2, 2026
Repositorymukul975/anthropic-cybersecurity-skills

What it does

Run a structured investigation when logs or alerts surface suspicious hashes, IPs, domains, or URLs.

Files

SKILL.mdMarkdownGitHub ↗

Analyzing Indicators of Compromise

When to Use

Use this skill when:

  • A phishing email or alert generates IOCs (URLs, IP addresses, file hashes) requiring rapid triage
  • Automated feeds deliver bulk IOCs that need confidence scoring before ingestion into blocking controls
  • An incident investigation requires contextual enrichment of observed network artifacts

Do not use this skill in isolation for high-stakes blocking decisions — always combine automated enrichment with analyst judgment, especially for shared infrastructure (CDNs, cloud providers).

Prerequisites

  • VirusTotal API key (free or Enterprise) for multi-AV and sandbox lookup
  • AbuseIPDB API key for IP reputation checks
  • MISP instance or TIP for cross-referencing against known campaigns
  • Python with requests and vt-py libraries, or SOAR platform with pre-built connectors

Workflow

Step 1: Normalize and Classify IOC Types

Before enriching, classify each IOC:

  • IPv4/IPv6 address: Check if RFC 1918 private (skip external enrichment), validate format
  • Domain/FQDN: Defang for safe handling (evil[.]com), extract registered domain via tldextract
  • URL: Extract domain + path separately; check for redirectors
  • File hash: Identify hash type (MD5/SHA-1/SHA-256); prefer SHA-256 for uniqueness
  • Email address: Split into domain (check MX/DMARC) and local part for pattern analysis

Defang IOCs in documentation (replace . with [.] and :// with [://]) to prevent accidental clicks.

Step 2: Multi-Source Enrichment

VirusTotal (file hash, URL, IP, domain):

import vt

client = vt.Client("YOUR_VT_API_KEY")

# File hash lookup
file_obj = client.get_object(f"/files/{sha256_hash}")
detections = file_obj.last_analysis_stats
print(f"Malicious: {detections['malicious']}/{sum(detections.values())}")

# Domain analysis
domain_obj = client.get_object(f"/domains/{domain}")
print(domain_obj.last_analysis_stats)
print(domain_obj.reputation)
client.close()

AbuseIPDB (IP addresses):

import requests

response = requests.get(
    "https://api.abuseipdb.com/api/v2/check",
    headers={"Key": "YOUR_KEY", "Accept": "application/json"},
    params={"ipAddress": "1.2.3.4", "maxAgeInDays": 90}
)
data = response.json()["data"]
print(f"Confidence: {data['abuseConfidenceScore']}%, Reports: {data['totalReports']}")

MalwareBazaar (file hashes):

response = requests.post(
    "https://mb-api.abuse.ch/api/v1/",
    data={"query": "get_info", "hash": sha256_hash}
)
result = response.json()
if result["query_status"] == "ok":
    print(result["data"][0]["tags"], result["data"][0]["signature"])

Step 3: Contextualize with Campaign Attribution

Query MISP for existing events matching the IOC:

from pymisp import PyMISP

misp = PyMISP("https://misp.example.com", "API_KEY")
results = misp.search(value="evil-domain.com", type_attribute="domain")
for event in results:
    print(event["Event"]["info"], event["Event"]["threat_level_id"])

Check Shodan for IP context (hosting provider, open ports, banners) to identify if the IP belongs to bulletproof hosting or a legitimate cloud provider (false positive risk).

Step 4: Assign Confidence Score and Disposition

Apply a tiered decision framework:

  • Block (High Confidence ≥ 70%): ≥15 AV detections on VT, AbuseIPDB score ≥70, matches known malware family or campaign
  • Monitor/Alert (Medium 40–69%): 5–14 AV detections, moderate AbuseIPDB score, no campaign attribution
  • Whitelist/Investigate (Low <40%): ≤4 AV detections, no abuse reports, legitimate service (Google, Cloudflare CDN IPs)
  • False Positive: Legitimate business service incorrectly flagged; document and exclude from future alerts

Step 5: Document and Distribute

Record findings in TIP/MISP with:

  • All enrichment data collected (timestamps, source, score)
  • Disposition decision and rationale
  • Blocking actions taken (firewall, proxy, DNS sinkhole)
  • Related incident ticket number

Export to STIX indicator object with confidence field set appropriately.

Key Concepts

TermDefinition
IOCIndicator of Compromise — observable network or host artifact indicating potential compromise
EnrichmentProcess of adding contextual data to a raw IOC from multiple intelligence sources
DefangingModifying IOCs (replacing . with [.]) to prevent accidental activation in documentation
False Positive RatePercentage of benign artifacts incorrectly flagged as malicious; critical for tuning block thresholds
SinkholeDNS server redirecting malicious domain lookups to a benign IP for detection without blocking traffic entirely
TTLTime-to-live for an IOC in blocking controls; IP indicators should expire after 30 days, domains after 90 days

Tools & Systems

  • VirusTotal: Multi-engine malware scanner and threat intelligence platform with 70+ AV engines, sandbox reports, and community comments
  • AbuseIPDB: Community-maintained IP reputation database with 90-day abuse report history
  • MalwareBazaar (abuse.ch): Free malware hash repository with YARA rule associations and malware family tagging
  • URLScan.io: Free URL analysis service that captures screenshots, DOM, and network requests for phishing URL triage
  • Shodan: Internet-wide scan data providing hosting provider, open ports, and banner information for IP enrichment

Common Pitfalls

  • Blocking shared infrastructure: CDN IPs (Cloudflare 104.21.x.x, AWS CloudFront) may legitimately host malicious content but blocking the IP disrupts thousands of legitimate sites.
  • VT score obsession: Low VT detection count does not mean benign — zero-day malware and custom APT tools often score 0 initially. Check sandbox behavior, MISP, and passive DNS.
  • Missing defanging: Pasting live IOCs in emails or Confluence docs can trigger automated URL scanners or phishing tools.
  • No expiration policy: IOCs without TTLs accumulate in blocklists indefinitely, generating false positives as infrastructure is repurposed by legitimate users.
  • Over-relying on single source: VirusTotal aggregates AV opinions — all may be wrong or lag behind emerging malware. Use 3+ independent sources for high-stakes decisions.

Related skills

FAQ

Is Analyzing Indicators Of Compromise safe to install?

skills.sh reports 1 of 3 security scanners passed. Review the Security Audits panel on this page before installing in production.

Securityauditappsec

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.