Got 5 alerts in the queue, all phishing-related. Mix of inbound email alerts and one firewall block. Went through each one to figure out what's real and what's noise.
| Alert ID | Rule | Verdict |
|---|---|---|
| #8814 | Inbound Email — Suspicious External Link | False Positive |
| #8815 | Inbound Email — Suspicious External Link | False Positive |
| #8816 | Access to Blacklisted URL — Blocked by Firewall | True Positive |
| #8817 | Inbound Email — Suspicious External Link | True Positive |
| #8818 | Inbound Email — Suspicious External Link | True Positive |
Both were legitimate emails that happened to contain external links. #8814 was an HR
onboarding email sent to a new employee (j.garcia), linking to hrconnex.thm —
clean domain, no endpoint activity in proxy logs. #8815 was similar. The detection rule is
too broad; it flags anything with an external link regardless of context.
These were the real ones. #8817 was a fake Amazon delivery email pushing the recipient to click a bit.ly link under a 48-hour urgency threat. Sender domain didn't match Amazon's actual infrastructure. Ran the link through VirusTotal — flagged as phishing. Confirmed in Splunk proxy logs that the user actually clicked it:
index=main sourcetype=proxy dest_domain="bit.ly" | table _time, src_ip, dest_url, action
#8818 was the same pattern, different sender spoof.
This one made sense once I connected it to #8817. The user clicked the bit.ly link; the firewall caught it and blocked the outbound connection. So #8816 wasn't a standalone event — it was the follow-on from the phishing email being clicked. Always worth correlating firewall blocks with email alerts around the same timestamp.
| Type | Value |
|---|---|
| URL | http://bit.ly/3sHkX3da12340 |
| Sender | Spoofed Amazon look-alike domain |
| Destination | Phishing site (flagged on VirusTotal) |
T1566.002 — Spearphishing Link T1036 — Masquerading (spoofed sender)
The main lesson here was not treating all 5 alerts the same. Two were noisy rules, three were an actual attack chain. The firewall block (#8816) only made sense after correlating it with the email timestamps — that's the kind of pivoting that matters in real triage.
bit.ly links in emails are almost always a red flag. Never trust a shortened URL in an unsolicited email.