GitHub Trends
Project Radar
- Blue team toolsublime-security/sublime-rulesYAML · 368 stars
- Dual-use project0xDanielLopez/TweetFeedRepo · 671 stars
- Dual-use projectphishdestroy/destroylistHTML · 1,662 stars
Research DeskLatest update July 19, 202659 research entries
Independent Research Desk
Phishing Tradecraft · Infrastructure · Detection Engineering
PhishPond researches how modern phishing operations are built, run, and detected — campaign evolution, adversary infrastructure, phishing kits, OAuth and device-code abuse, AiTM frameworks, and the detection and validation workflows that catch them.
Recurring Intel
Attack-Side Tradecraft
Campaign tradecraft, lure mechanics, adversary infrastructure, identity pressure, and operator workflows worth modeling.
12 attack-side readsDetection Engineering
Detection engineering, telemetry analysis, reporting workflows, and validation that security teams can operationalize.
32 detection readsAPT Tradecraft
Emerging procedures, tooling, initial-access patterns, and cross-team tradecraft from real-world actor reporting.
15 tradecraft readsGitHub Trends
New Today
Lead Research
The authentication failures look fragmented by app, but the campaign becomes visible when defenders pivot on missing app names, error semantics, source behavior, and tenant-wide volume.
Two large Entra ID campaigns used hundreds of thousands to millions of fictional OAuth client IDs to spread account enumeration across apparent applications.
Live Collection
Fake verification steps train users into running attacker-provided instructions.
Read more:The Hacker News
A successful click delivers persistent remote access, not just credentials.
Read more:The Hacker News
Trusted suppliers and developer channels can carry phishing risk past normal filters.
Read more:BleepingComputer
Fake verification steps train users into running attacker-provided instructions.
Read more:BleepingComputer
Fake verification steps train users into running attacker-provided instructions.
Read more:The Hacker News
Sublime rules for email attack detection, prevention, and threat hunting. Primary language: YAML. 368 stars.
Open project:GitHub
#email-security#phishing#threat-hunting
TweetFeed collects Indicators of Compromise (IOCs) shared by the infosec community at Twitter. Here you will find malicious URLs, domains, IPs, and SHA256/MD5 hashes. 671 stars.
Open project:GitHub
#blueteam#malware#malware-detection#malware-research
Real-time phishing & scam domain blocklist — 200k+ curated threats, 888K+ community, free API, multiple formats Primary language: HTML. 1,662 stars.
Open project:GitHub
#anti-phishing#blacklist#blocklist#crypto-scam
🐟 PhishTank Blocklist for Pi-hole Primary language: Shell. 12 stars.
Open project:GitHub
#blocklist#hosts#phishing#pihole
Aggregation of lists of malicious domains (phishing) that can be integrated into FortiGate firewalls and other products. Primary language: DIGITAL Command Language. 110 stars.
Open project:GitHub
#blocklist#blocklists#domains-blacklist#domains-list
Coverage Map
Specific campaigns, actor activity, and the lures behind them.
How techniques work end-to-end — walkthroughs and operator workflows.
Adversary infrastructure: kits, AiTM, redirectors, and sending abuse.
Detection engineering, telemetry, validation, and response.
Longer research notes, measurement, and periodic briefs.
Search Tool
Search titles, authors, tags, and body text across the PhishPond research archive.
Showing 9 matching entries.Clear search
Field Analysis
Phishing kits get the headlines, but the hosting underneath them is the durable asset. A May 2026 seizure of 800+ servers, resilient scanning networks, and the routine hop behind a CDN show why takedowns rarely stick and where the defensible signal actually lives.
Read more:CISA (with NSA, DC3, FBI, and international partners)ELLIO
Field Analysis
A fake recruiter asking a candidate to review an MVP repo shows why unsolicited source code is not a document. It is an executable threat surface with access to developer secrets.
Field Analysis
Socket attributes a coordinated supply-chain campaign called TrapDoor to roughly thirty-four packages across npm, PyPI, and Crates.io, with ecosystem-specific execution paths and a new twist: planted .cursorrules and CLAUDE.md files designed to influence the developer's AI coding assistant.
Read more:SocketThe Hacker News
Field Analysis
A reported exploitation wave against Ghost CMS pushed malicious JavaScript onto more than 700 sites, sending visitors into fake verification flows that used ClickFix-style paste-and-run instructions.
Read more:The Hacker NewsMalwarebytes Labs
Field Analysis
GitHub's staged publishing and new npm install-source controls give maintainers practical ways to slow compromised CI/CD paths before a malicious package becomes installable.
Read more:GitHub ChangelogCISA
Field Analysis
Recent actor reporting points to a practical trend line: adversaries are combining selective delivery, user-driven execution, and trusted developer channels.
Read more:The Hacker NewsDark Reading
Field Analysis
The Salesloft Drift incident showed how a trusted integration token can become an access path into customer SaaS data without a fresh user login.
Read more:The Hacker NewsThe Hacker News
Field Analysis
Abuse of legitimate email services such as Amazon SES shows why authentication pass results are not the same thing as sender trust.
Read more:BleepingComputerMicrosoft Security Blog
Field Analysis
Recent package compromises show how developer trust can be abused to harvest credentials and seed downstream phishing risk.
Read more:BleepingComputerCISA