Analyzing Email Headers For Phishing Investigation

Analyzing Email Headers for Phishing Investigation

When to Use

• When investigating a suspected phishing email to determine its true origin
• For verifying sender authenticity and detecting email spoofing
• During incident response when a user has clicked a phishing link
• When tracing the delivery path and relay servers of a suspicious email
• For validating SPF, DKIM, and DMARC alignment to identify forgery

Prerequisites

• Raw email headers from the suspicious message (EML or MSG format)
• Understanding of SMTP protocol and email header fields
• Access to DNS lookup tools (dig, nslookup) for SPF/DKIM/DMARC verification
• Email header analysis tools (MHA, emailheaders.net concepts)
• Python with email parsing libraries for automated analysis
• Access to threat intelligence platforms for IP/domain reputation

Workflow

Step 1: Extract Raw Email Headers

# Export from Outlook: Open email > File > Properties > Internet Headers
# Export from Gmail: Open email > Three dots > Show original
# Export from Thunderbird: View > Message Source

# If working with EML file from forensic image
cp /mnt/evidence/Users/suspect/AppData/Local/Microsoft/Outlook/phishing_email.eml \
   /cases/case-2024-001/email/

# If working with PST file, extract individual messages
pip install pypff
python3  0.8:
    print("WARNING: Likely typosquatting/lookalike domain!")
PYEOF

# Check domain reputation on VirusTotal
curl -s "https://www.virustotal.com/api/v3/domains/${SENDER_DOMAIN}" \
   -H "x-apikey: YOUR_VT_API_KEY" | python3 -m json.tool

# Check if the Reply-To differs from From (common phishing indicator)
python3 -c "
import email
with open('/cases/case-2024-001/email/phishing_email.eml') as f:
    msg = email.message_from_file(f)
from_addr = email.utils.parseaddr(msg['From'])[1]
reply_to = email.utils.parseaddr(msg.get('Reply-To', msg['From']))[1]
if from_addr != reply_to:
    print(f'WARNING: From ({from_addr}) != Reply-To ({reply_to})')
else:
    print('From and Reply-To match')
"

Step 5: Examine Email Body and Attachments

# Extract URLs from email body
python3 "\']+', content)
    print("=== URLs FOUND IN EMAIL BODY ===")
    for url in set(urls):
        print(f"  {url}")

    # Check for URL obfuscation (display text != href)
    href_pattern = re.findall(r']*href=["\']([^"\']+)["\'][^>]*>(.*?)', content, re.DOTALL)
    print("\n=== HYPERLINK ANALYSIS ===")
    for href, text in href_pattern:
        display_url = re.findall(r'https?://[^\s Actual='{href}'")

# Extract and hash attachments
print("\n=== ATTACHMENTS ===")
for part in msg.walk():
    if part.get_content_disposition() == 'attachment':
        filename = part.get_filename()
        content = part.get_payload(decode=True)
        import hashlib
        sha256 = hashlib.sha256(content).hexdigest()
        print(f"  File: {filename}, Size: {len(content)}, SHA-256: {sha256}")
        with open(f'/cases/case-2024-001/email/attachments/{filename}', 'wb') as af:
            af.write(content)
PYEOF

# Submit attachment hashes to VirusTotal
# Submit URLs to URLhaus or PhishTank for reputation check

Key Concepts

| Concept | Description | |---------|-------------| | SPF (Sender Policy Framework) | DNS record specifying authorized mail servers for a domain | | DKIM (DomainKeys Identified Mail) | Cryptographic signature verifying email content integrity | | DMARC | Policy framework combining SPF and DKIM for sender authentication | | Received headers | Server-added headers showing each hop in the delivery chain (read bottom to top) | | Return-Path | Envelope sender address used for bounce messages; may differ from From | | Message-ID | Unique identifier assigned by the originating mail server | | X-Originating-IP | Original sender IP address (added by some mail services) | | Header forgery | Attackers can forge From, Reply-To, and other headers but not Received chains |

Tools & Systems

| Tool | Purpose | |------|---------| | MXToolbox | Online email header analyzer and DNS lookup | | dig/nslookup | DNS record queries for SPF, DKIM, DMARC verification | | pyspf | Python SPF record validation library | | dkimpy | Python DKIM signature verification library | | PhishTool | Specialized phishing email analysis platform | | VirusTotal | URL and file reputation checking service | | AbuseIPDB | IP address reputation database | | whois | Domain registration information lookup |

Common Scenarios

Scenario 1: CEO Fraud / Business Email Compromise The email claims to be from the CEO but Reply-To points to a Gmail address, SPF fails because the sending IP is not authorized for the spoofed domain, DKIM is missing, and the From domain is a lookalike (ceo-company.com vs company.com).

Scenario 2: Credential Harvesting Phishing Email contains a link that displays "login.microsoft.com" but href points to a lookalike domain, the attachment is an HTML file containing a fake login page with credential exfiltration JavaScript, the sending domain was registered 3 days ago.

Scenario 3: Malware Delivery via Attachment Email with an Office document attachment containing macros, the sender domain passes SPF but the account was compromised, DKIM signature is valid (sent from legitimate infrastructure), attachment SHA-256 matches known malware on VirusTotal.

Scenario 4: Spear Phishing with Legitimate Service Attacker uses a legitimate email marketing service to send phishing, SPF and DKIM pass because the service is authorized, the phishing is in the content not the infrastructure, requires URL and content analysis rather than header authentication checks.

Output Format

Email Header Analysis Report:
  Subject:     "Urgent: Invoice Payment Required"
  From:        accounting@examp1e-corp.com (SPOOFED)
  Reply-To:    payments.urgent@gmail.com (MISMATCH)
  Return-Path: 
  Date:        2024-01-15 09:23:45 UTC

  Delivery Path (4 hops):
    Hop 1: mail-server.xyz [203.0.113.45] -> relay1.isp.com
    Hop 2: relay1.isp.com -> mx.target-company.com
    Hop 3: mx.target-company.com -> internal-filter.target.com
    Hop 4: internal-filter.target.com -> mailbox

  Authentication:
    SPF:    FAIL (203.0.113.45 not authorized for examp1e-corp.com)
    DKIM:   NONE (no signature present)
    DMARC:  FAIL (p=none, no enforcement)

  Indicators of Phishing:
    - Lookalike domain (examp1e-corp.com vs example-corp.com, 96% similar)
    - From/Reply-To mismatch
    - Domain registered 2 days before email sent
    - URL in body points to credential harvesting page
    - Attachment: invoice.xlsm (SHA-256: a3f2...) - Known malware on VT

  Risk Level: HIGH

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: mukul975
• Source: mukul975/Anthropic-Cybersecurity-Skills
• License: Apache-2.0
• Homepage: https://mahipal.engineer/Anthropic-Cybersecurity-Skills/

Install and usage instructions live in the source repository linked above.