惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

The GitHub Blog
The GitHub Blog
Martin Fowler
Martin Fowler
Vercel News
Vercel News
U
Unit 42
Engineering at Meta
Engineering at Meta
aimingoo的专栏
aimingoo的专栏
MyScale Blog
MyScale Blog
Y
Y Combinator Blog
阮一峰的网络日志
阮一峰的网络日志
爱范儿
爱范儿
Apple Machine Learning Research
Apple Machine Learning Research
博客园_首页
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
B
Blog RSS Feed
N
Netflix TechBlog - Medium
GbyAI
GbyAI
F
Fortinet All Blogs
MongoDB | Blog
MongoDB | Blog
大猫的无限游戏
大猫的无限游戏
C
Check Point Blog
M
MIT News - Artificial intelligence
D
Docker
IT之家
IT之家
Stack Overflow Blog
Stack Overflow Blog

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
We Replayed the Nomad Hack. Our Tool Fired at Block 15259...
Brett Moore · 2026-06-09 · via DEV Community

In August 2022, the Nomad bridge was drained of $190 million over eight hours. The on-chain signal was there from the very first transaction. Nobody caught it in time.

We built Heimdall to change that.

The Problem With Bridge Security

Cross-chain bridges are the most hacked infrastructure in crypto. Ronin: $625M. Wormhole: $320M. Nomad: $190M. The pattern repeats because bridges hold enormous amounts of locked value and monitoring tooling has never kept up.

Existing tools like Forta and OpenZeppelin Defender are general-purpose. They weren't built with bridge-specific heuristics in mind. The result: slow alerts, high false-positive rates, and teams flying blind during the most critical minutes of an exploit.

Heimdall is a focused, open-source bridge monitoring and anomaly detection system built specifically for cross-chain bridges. It watches TVL, mint/burn flows, and transaction patterns in real time and fires alerts when something looks wrong.


Three Rules, Three Signals

Heimdall's detection engine is rule-based by design. No black box, no ML, just precise heuristics that map directly to known exploit patterns.

The three core rules:

large-fill — fires when a single withdrawal exceeds a configurable % of TVL. Catches the opening move of most bridge drains.

rapid-drain — fires when N large fills occur within a rolling time window. Catches the copycat cascade that followed Nomad's initial exploit.

imbalance — fires when fills across all chains exceed deposits by a configurable multiplier. The most robust signal: legitimate bridge activity always has deposits on one side. A drain doesn't.


Replaying the Nomad Hack

To validate the detection logic, We replayed the Nomad exploit against real on-chain data.

Setup:

  • Block range: 15259000 → 15261500 (~8 hour window)
  • Pre-seeded TVL matching Nomad's actual holdings: $85M USDC, $40M DAI, ~$26.5M WETH, ~$30M WBTC
  • 1,308 real events indexed from RPC, 1,208 of which were fills

Results:

[CRITICAL][large-fill]   Single fill of 33.3% of TVL on nomad chain 1
                         @ block 15259101 (+0.0 min into hack)

[CRITICAL][rapid-drain]  20 fills in a 50-block window on nomad
                         for token 0x2260fac5… (WBTC)

[HIGH][imbalance]        WBTC fills exceed deposits by 2,570,626.8%
                         across all chains on nomad

First CRITICAL alert: block 15259101, the same block the exploit began.

Not ten minutes later. Not after the second wave of copycats. The first transaction.

The large-fill rule caught a single fill representing 33% of Nomad's total TVL. The imbalance rule flagged a 2.5 million percent excess of fills over deposits, a number that is simply impossible under any legitimate bridge activity.

Had Heimdall been running on August 1, 2022, it would have fired before a single copycat joined the drain.


What the Numbers Mean

The 2,570,626% imbalance figure deserves a moment. That's not a threshold tuning question. That's not a noisy signal requiring human judgment. That is an unambiguous on-chain fact: millions of dollars leaving a bridge with nothing coming in.

The Nomad hack was unique because it was chaotic, nearly 300 addresses piled on over 150 minutes, many just copying the original transaction calldata. That chaos is exactly what the rapid-drain rule was built to catch: not one sophisticated actor, but a flood of sequential large withdrawals in a tight window.


What's Next

  • Real-time Telegram alerts
  • Public dashboard with live TVL and alert feed
  • Retroactive analysis of the Hop Protocol exploit
  • Historical false-positive audit against 6 weeks of Across data The repo is live at github.com/moorebrett0/heimdall. If you work on a bridge team and want your bridge added, open an issue it's just a config file, not a PR into core logic.

The Broader Point

Bridges are not going away. Cross-chain activity is growing. The tooling to protect it needs to catch up.

Heimdall is one piece of that. Open source, focused, and built on the premise that the signal for most bridge exploits is not subtle, it's just not being watched.

Block 15259101 was eight hours before the last dollar left Nomad. That's eight hours to alert, respond, pause, and potentially save $190 million.

That window exists for the next bridge too. Heimdall is watching for it.