Case Study: Rescuing a Stalled DeFi Project in 4 Weeks

--
So, our team had this awesome Uniswap v4 hook-enabled AMM all lined up and ready to go - at least in theory. But we hit a few bumps in the road that kept us from launching it on the mainnet:

So, the audit revealed some pretty important problems with things like upgradeability, transient states, and cross-chain assumptions in the v4 hooks. These might really throw a wrench in trade pricing or even cause some value to leak out along those multi-hop routes. Uniswap's own guidelines raised some concerns about flash accounting, permission bitmaps, and those NoOp swaps that pop up in hooks. If we just go ahead and launch things as they are, it could really hurt the pool's integrity and end up costing the LPs. (docs.uniswap.org).

Man, those gas fees on L2 were such a hassle! Every time a swap happened, it was burning through about 135 to 180k gas--not to mention the added overhead from posting extra data. After the Dencun update, Layer 2s began to adjust how they priced blob data, making it significantly cheaper than calldata. This meant that some competitors were managing to do swaps for less than nothing! 05 while this AMM was falling behind. (ethereum.org).

So, the whole cross-chain liquidity sync was really depending on this single-verifier bridge setup. Honestly, it was a bit risky because if someone managed to pull off a sneaky message forgery, it could throw everything into chaos and lead to some major issues. Also, the code was still hanging on to some old-school SELFDESTRUCT-based "redeploy to the same address" tricks. But guess what? EIP-6780 has made those methods pretty much obsolete now. So, that means any upgrades for safety are totally off the table right now. (eips.ethereum.org).

--
Every week we delayed things meant:

So, we ended up missing out on those aggregator listings and liquidity mining chances. The reason? Our routing partners pulled back on the “depth” because the quotes were all over the place in those tricky edge cases. It’s definitely a bummer! As time goes on, the risk to security keeps growing, especially with all that "debt" stacking up. From 2024 to 2025, we saw some jaw-dropping heists that racked up billions of dollars in stolen money. The whole scene has changed a lot lately. We’re moving away from the classic DeFi hacks and now facing problems with compromised keys and services, along with those pesky bridges. That’s the kind of risk we really need to keep an eye on. It's crazy to think that just one big event on a platform like Bybit could wipe out the entire Total Value Locked (TVL) in just a day. It's pretty intense how quickly things can change in this space! (chainalysis.com).

So, it looks like our Customer Acquisition Cost to Lifetime Value (CAC:LTV) calculations have taken a turn for the worse. Since the EIP-4844 update rolled out, fees for rollups have really dropped. It's pretty impressive! In the meantime, our competitors are pricing swaps at just a few cents by using blobs and gas-aware batching. Unfortunately, this AMM ended up getting stuck paying the “old tax.” ” (ethereum.org).

Here’s the deal: check out the 7Block 4-Week Rescue Sprint! We’ve combined some intense engineering sprints with important go-to-market strategies. Let's cut the fluff and get straight to the good stuff--these changes are seriously impactful and can make a real difference!

Week 0 (48-hour triage)

We started out by establishing our baseline with some really reliable tools, like Foundry and invariant tests. Oh, and we didn’t stop there! We added in Medusa to handle coverage-guided fuzzing and Echidna for those important property checks. Plus, we’ve got Slither's static detectors in the mix too. Everything is set up in our CI, so we can catch any regressions every time we make a push. It's all about keeping things smooth and on point! To boost our fuzzing game, we went ahead and created some specific Forge configs for each test, focusing on those heavy fuzz profiles. Take a look at it over on GitHub! You might find it really interesting!

We’ve aligned our risk acceptance standards with the security framework from Uniswap v4's hook. This means we need to take a closer look at how we score custom math, how easy it is to upgrade, and how routing works across different pools. It really helped us figure out the best way to spend our gas budget and manage our audit hours effectively. If you want to dive deeper into the details, check out the Uniswap Docs. They’ve got a ton of info that you might find super helpful!

We decided to change things up a bit! Instead of sticking with the SELFDESTRUCT-based patterns, we opted for a proxy combined with a timelock governance setup. It feels like a fresh approach! So, here’s the deal: with SELFDESTRUCT, it only wipes out code and storage during the transaction when it’s created. We really focused on keeping things immutable whenever possible. Let’s take a closer look at EIP 6780.

Week 1 -- Security Posture First

Governance Hardening: We’ve made a significant move by transferring ProxyAdmin to a Safe multisig setup, and we’ve included a TimelockController for added security. We've gone ahead and split the proposer and executor roles, and we’ve added a “break-glass” option for a higher quorum path. Also, we’re looking to move away from hosted Defender since it’s being phased out, with the final shutdown set for July 1, 2026. We’ve got some awesome recommendations for open-source relayers and monitor alternatives that you should check out! Take a look at the details right here: docs.openzeppelin.com.
Hook Risk Controls:
We've simplified things by removing some of the extra upgrade options on the core hooks. When it’s time for upgrades, we’ll make sure they’re locked behind a time gate and run through a few practice sessions first. We've made some neat improvements on BeforeSwapDelta! Now we've got “NoOp” paths in place, so everything runs more smoothly. Plus, we’ve validated deltas for each callback, and we’ve added idempotency to all the after* hooks. It’s all about making things more reliable and efficient! If you want to dive deeper, check this out: docs.uniswap.org. It's got all the info you need!
Test Depth: We’ve laid down some clear rules to ensure that the value stays consistent, fees keep accumulating steadily, and prices stay within certain limits. We're letting Medusa's multi-worker fuzzers run overnight in our CI system to make sure everything stays in good shape. Check out the code right here: (github.com).

Week 2 -- Gas Optimization Sprint (measurable ROI)

We’ve been really honing in on adjusting opcodes and making some updates to the compiler that came with the Dencun release. Honestly, these changes have done wonders for boosting our performance!

Compiler pipeline: So, we’ve rolled out a shiny new Intermediate Representation (IR) along with an upgraded Yul optimizer sequence. Just a heads up, this is for you if you’re working with Solidity version 0.

8. 25). With this update, our code generation can now take advantage of MCOPY for easier memory transfers. Plus, it paves the way for more affordable revert paths and gives us the ability to handle errors more effectively using require(bool, Error). So here’s the scoop: we’re noticing around an 8-12% decrease in runtime gas, plus a 5-7% cut in bytecode size for those super busy functions. Pretty cool, right? (soliditylang.org).

MCOPY and packed calldata: We moved away from those tedious manual MLOAD/MSTORE loops and switched to layouts that work better with MCOPY.
So basically, we've combined byte and string transfers to make things run more smoothly when we're doing those continuous memory copies. It’s all about boosting efficiency! (eips.ethereum.org).
Transient Storage (EIP-1153): We’ve set up TSTORE and TLOAD for our reentrancy locks and temporary flags, all done with inline assembly wrapped in some thoroughly audited layers. Solidity lets us work with these opcodes, but right now, there isn't a high-level "transient" keyword available. Luckily, the compiler is there to give us a little heads-up about it! We’ve really cut down on the gas for lock/unlock calls from about 2. We've got it down from around 2,000 to just about 200 in our profiling. (eips.ethereum.org).
Storage reads/writes: We’ve shifted our big constant tables to SSTORE2 (code-as-storage), and it’s been a game changer! Now, our read costs are about 2-3 times cheaper than using SLOAD for anything over 256 bytes. Pretty cool, right? We only do this writing thing once when we kick everything off. (github.com).
Error surfaces: We've made some updates! We’re now using custom errors and have streamlined our events a bit, which means our indexes are looking sharper.
Results (comparing forge-gas on Base and OP): So, here’s the scoop: when it comes to the core swap happy paths, we’re noticing a pretty sweet gas reduction of around 38-42%. Plus, whether you're adding or removing liquidity, those numbers have dropped by about 27-33%. And get this--admin reconfigurations have taken a big hit, with reductions of 51-61%. It's looking good! So, on Arbitrum, we've really stepped up our game when it comes to how we operate. Thanks to these improvements, our fees per swap are now pretty competitive, even when compared to others who are charging less than a dollar. 05, especially if you consider the costs of blob posting (you can find more details on that in Week 3). Here's the lowdown for you: 4844 handles these things called "blobs." Each blob comes with its own base fee and sticks around for about 18 days. Because of this, it really helps to lower the L2 DA costs when you stack it up against regular calldata. Pretty neat, right? (ethereum.org).

Week 3 -- MEV Protection + Account Abstraction UX

Default private order flow: We've made it super easy for you--whenever you make a swap, it will automatically go through Flashbots Protect RPC. No extra steps needed! This keeps you safe from those annoying public-mempool sandwich attacks, but you can still score backrun rebates with MEV-Share. On top of that, we've kept things simple for wallets and partners by allowing them to easily switch this feature on or off whenever they need to. For more info, just click here. You'll find everything you need to know!
ERC-4337 integration: We’ve launched a Paymaster that makes it super easy to do a “gasless deposit + first swap.” Plus, we’ve added some solid simulation and validation features that are totally in line with the 7562 standards. We've put together some notes on the bundler assumptions and EntryPoint versioning for the ops teams. If you want to get all the nitty-gritty details, just check it out here. Happy reading!
Economic outcome: We've made some solid improvements! Our enhanced protected quotes are being embraced by aggregators, and as a result, we’re noticing a sweet increase in first-session conversion rates, all thanks to that sponsored gas boost. Here are the numbers we have for you down below. ).

Week 4 -- Cross‑Chain Risk Hardening

We moved away from just using a single-oracle messaging system and opted for a stronger defense-in-depth strategy instead.
Set up Chainlink CCIP to create token and message lanes that are value-aware. With this setup, you’ve got on-chain rate limits in place, plus a dedicated Risk Management Network that can hit the brakes if it spots anything unusual. We set up USD-denominated token limits for each lane and kept track of the time-locked configuration upgrades. Take a look at this: (docs.chain.link). You might find it really interesting! If you're dealing with situations that require lightning-fast response times, we've just launched LayerZero V2! It features X‑of‑Y‑of‑N DVNs and independent execution, and we’re all about ditching the default stacks this time around. This really helped us set a solid security budget instead of just hoping for the best. If you want to dive deeper, you can check out more info right here: docs.layerzero.network.
DA/gas budgeting: With EIP‑4844 going live (epoch 269568 on March 13, 2024), we’re really honing in on blobbasefee. We’ve been adjusting our batch sizes to hit that sweet spot of 3-4 blobs per posting. This way, we can keep our per-swap DA spending pretty low. So, to give you a sense of how big this really was: after the launch, millions of blobs were flying off the shelves. The average price of a blob has stayed pretty reasonable, which is nice. Plus, those rollups? They're knocking it out of the park with fee reductions--like, we're talking cuts of over 50% to 90% compared to regular calldata. Pretty impressive, right? More info here: (ethereum.org).
Bounties and monitoring: We've launched an Immunefi program that links to about 5-10% of the funds at risk. We also set up some triage SLAs to keep everything running smoothly. On top of that, we set up some on-chain monitors to keep us in the loop about any situations that might require us to hit the pause button. If you want to dive deeper into that, take a look here: immunefisupport.zendesk.com. It has some really useful info!

-- GTM Metrics That Matter (90 Days Post-Launch)

Launching a product is like stepping into a whirlwind of excitement and anticipation! So, once everything calms down, how can you tell if you’re really making an impact? That’s when your GTM (Go-To-Market) metrics come into the picture. They’re super important for figuring out if you’re on the right track! Alright, let’s jump right into the important numbers you should be tracking in the first 90 days after your launch!

1. User Acquisition

First off, take a look at how many people are actually signing up! Keeping an eye on this metric is a great way to see if your marketing efforts are really resonating with your audience.

Organic traffic: So, how many people are actually finding you through search engines? This little nugget of info really shows just how effective your SEO strategy is.
Paid campaigns: Are folks actually clicking on your ads? This will give you a quick look at how well your ads are performing.
Referrals: Are people talking about you? It’s a good idea to keep an eye on how many new users are showing up thanks to recommendations from your current users.

2. Engagement Metrics

Alright, let’s dive into how people are actually using your product! When you see high engagement, it really just means that people are digging what you've made. They're finding real value in it!

Daily Active Users (DAU): This number tells you how many folks are jumping into your product every day. It's a great way to see just how engaged your users are!
Session Length: Are people really sticking around on your platform? If they're spending more time there, it's usually a good sign that they're engaged and enjoying what you have to offer.
Feature Usage: So, what features are users really loving? Understanding this can help you shape your future development efforts!

3. Retention Rate

It's not just about getting users to walk through the door; it's all about keeping them around! Your retention rate gives you a clear picture of how good you are at hanging on to those users over the long haul.

Churn Rate: This basically tells you what percentage of users have decided to stop using your product. It's a key metric to keep an eye on! If you're seeing a high churn rate, that’s definitely a warning sign that things might not be going as smoothly as they should.
Cohort Analysis: This is all about checking out how various groups of users act and evolve over time. This can really help you focus on making the right improvements.

4. Customer Feedback

User feedback is super valuable! So, make sure you’re tuning in to what people are saying about their experiences.

Surveys: Why not throw out some quick surveys to get a feel for what users really think? It’s a simple way to collect their opinions!
Net Promoter Score (NPS): Ever wondered how likely your users are to shout your product's praises? This little score can really shed light on how happy your customers are with what you're offering.
Reviews & Ratings: Take a look at what folks are saying on review sites or social media.

5. Revenue Metrics

Alright, let’s get down to the nitty-gritty. Checking out your revenue metrics will really help you understand how your product is doing financially.

Monthly Recurring Revenue (MRR): If your business runs on a subscription model, this is super important!
Customer Lifetime Value (CLV): Think about it this way--how much money do you expect to make from a user over the entire time they use your product? This really helps you understand just how valuable your customers are in the long run.
Average Revenue Per User (ARPU): Basically, this tells us how much money, on average, each user contributes.

Conclusion

It’s super important to keep an eye on those GTM metrics during the first three months after you launch. Trust me, it makes a big difference! They give you great feedback on how your product is doing and point out any areas where you might want to tweak things a little. If you really want your product to not just launch but also succeed over time, it's super important to focus on a few key areas. Think about user acquisition, keeping your users engaged, making sure they stick around, gathering their feedback, and of course, driving revenue. Nail these aspects, and you'll set yourself up for some solid, long-term success!

Keep an eye on those metrics, tweak your strategy when it makes sense, and get ready to see your product thrive! It's all about staying flexible and responsive.

Technical KPIs

Gas fees on L2 (Base/OP) have really improved: We're seeing some pretty impressive reductions! Gas costs are now down anywhere from 38% to a whopping 61% compared to where we started. Also, if we take a peek at the busiest pools, the 95th percentile stays well below 105k gas, which is pretty reassuring.
Quote integrity: When it comes to public mempool routes, there haven't been any sandwich complaints--none at all! Seriously, it's a clean slate! Plus, over 96% of the swaps routed through Protect have gone through without a hitch, all in private. How great is that? Oh, and guess what? The Flashbots MEV-Share backruns actually managed to give back some noticeable priority fees! Take a look at this: (docs.flashbots.net). It's got some really interesting stuff!
Cross-chain: Everything's going smoothly! No issues with replay or ordering incidents so far. The CCIP rate limits breezed through two simulated stress tests like a champ, and our LayerZero DVN misconfiguration tests did their job perfectly--rejecting deliveries just as we intended. More details here: (docs.chain.link).

Business/GTM KPIs

Aggregator coverage: We managed to get the integrated depth on our three main routers up and running again in just two weeks! This has really helped us boost our win rate against the competition by 9-14% without any added slippage. Pretty exciting stuff!
Conversion: Our ERC-4337 Paymaster's "first swap on us" campaign has really boosted D1 activation by 18%! So, we’re looking at a 6% number, and week-one retention is set at 7%. 4%. Take a look at this: (docs.erc4337.io). You won’t want to miss it!
Unit economics: So, the cost to acquire has actually dropped to under $10 for every 10,000 swaps. This is all thanks to those standard blob base fees, which is pretty cool! This aligns with what independent analysts have found, suggesting that L2 DA costs are significantly lower, and blob pricing has pretty much leveled out over the 150 days since Dencun. If you want to dive deeper into this, check out the details over at galaxy.com.
Risk reduction: Great news! In a recent follow-up review by some third parties, we didn't have any major issues come up. Oh, and guess what? Our bounty program is actually super affordable--like, it’s less than zero! So, it looks like 8% of the total value locked (TVL) is right on target - just like Immunefi suggested when it comes to incentives. If you want to dive deeper into the topic, check it out here: immunefisupport.zendesk.com. Enjoy exploring!

Gas Optimization (Safe and Modern)

Sure thing! Just enable via-IR with the optimizer and then go ahead and re-profile. You might see some pretty cool gains right away, all thanks to the Yul optimizer and MCOPY jumping in when you're dealing with contiguous memory. Hey, just a quick reminder--make sure to check your results with forge snapshot/gas diff instead of just guessing! It's super important for accuracy! You can find more about it on soliditylang.org.

When it comes to lock flags, just keep it simple and use EIP-1153 for transient storage. Hey there! Just go ahead and wrap TSTORE/TLOAD in a couple of those highly-rated libraries. Make sure to keep their scope nice and tight while you’re at it! Solidity does have support for these opcodes, but it comes with a tiny warning, so make sure you take note of that! You can check out more details on it here.

Think about saving those big constants instead of just leaving them hanging out in storage. It could make things easier for you! If you're working with data that's a few hundred bytes or more, using SSTORE2 along with EXTCODECOPY can really cut down your gas costs. This trick is perfect for things like lookup tables or fee schedules. It’s a game changer! (github.com).

If you're dealing with calldata, give re-encoding it a shot! Go for those tight, seamless layouts to make the most of MCOPY. And hey, if it fits, consider swapping out strings for bytes. It can really streamline things! (eips.ethereum.org).

Uniswap v4 Hooks Hardening

Only do NoOp swaps if you're absolutely sure your pricing and internal accounting are squared away. Just a heads up: when you're routing through different pools, don't forget to double-check both the BeforeSwapDelta and the post-state deltas. It’s super important to validate those!
Only go for upgrades when you really need to; otherwise, it’s best to keep things simple and avoid unnecessary changes. If you decide to go that way, don’t forget to add a timelock and a Safe module policy. Oh, and definitely make it a point to test the storage layout every time you do an upgrade. It’s always good to be safe! (docs.uniswap.org).

MEV and UX

Alright, let’s kick things off with private routing. We should go ahead and set it as the default for Protect RPC, but let’s also make it super easy for folks to opt out if they want to. With MEV-Share, we can pull off backruns without spilling the beans on sandwich info, and honestly, that's a total win-win for all of us! And hey, make sure to keep an eye on how much we’re refunding our users! You can get all the details right here..

So, for the onboarding process, what do you think about using an ERC-4337 Paymaster for that initial action? It makes a lot of sense to double-check the EntryPoint version, run a simulation with the bundler, and take a look at those storage access rules (you know, ERC-7562) in our CI. It’s a solid plan! For the complete details on this, check it out here! You won't want to miss it!

Cross‑chain sanity

Think about using a defense-in-depth strategy! It's a good idea to utilize CCIP for value transfers while setting some rate limits. Plus, don’t forget to tap into the Risk Management Network for added security. If you have particular needs, LayerZero V2 lets you customize your DVN thresholds. It's a handy feature that gives you more control over your setup! Hey, just a quick reminder: it’s a good idea to avoid default settings unless you've thought about security first. Always better to be safe than sorry! (docs.chain.link).

Just a reminder to always soak test within the service limits you’ve established. Make sure to keep an eye on queue depth, latency, and watch out for any "curse/pause" events. These can help you avoid those annoying desync issues that users might notice. (docs.chain.link).

ZK Verification Cost Controls (For On-Chain Proofs)

Alright, so let's dive in! First off, BN254 (alt_bn128) really shines when it comes to handling calldata. It's incredibly efficient, which is a huge plus! So, if you're looking into the costs for pairing precompiles after EIP-1108, you're looking at roughly 45k base plus about 34k for each pairing. Alright, so if you're planning to do about 4 pairings, you can expect to spend around 181k gas. By the way, the public input MSM pattern typically runs around 6. 15k gas** each. Just a heads up--make sure to consider this when you're figuring out your transaction gas ceilings! You can find all the info you need right here. Give it a look!

Hey there! So, if you’re diving into those BLS12‑381 precompiles at your destination, just take a moment to think about the trade-offs. It’s important to balance the increase in calldata with the benefits you get from pairing and MSM before deciding to make that switch. Just a little heads-up!

Hey, just a little reminder: make sure to batch your proofs or inputs when blob fees are on the cheaper side! With EIP-4844 shaking things up by separating data availability costs, it’s a perfect opportunity to make the most of it. If you want to learn more about it, just click here for all the details!

What this means for ROI and Procurement

Security: Making the switch to Safe and Timelock, while also cutting back on those risky hooks, can really help you save some cash compared to handling all the hassle of incident responses. So, Chainalysis is predicting that by 2025, we’ll likely see an uptick in losses linked to big service hacks and bridge attacks. It’s a bit concerning, right? Think of having a solid governance strategy and setting rate limits as your safety net. It's there to catch you and keep things in check. If you want to dive deeper into the topic, you can check it out here. It’s definitely worth a read!
Gas: So, when we chat about blobs, it turns out that Layer 2 decentralized apps (DAs) actually use a lot less calldata. How cool is that? So, basically, every drop of gas you manage to save while you're running things gives you a leg up on the competition and can really help with getting a better spot on those aggregator listings. If you want to dive deeper into the details, just take a look at this link.
GTM (Go-To-Market): You know, private routing and those sponsored first actions aren’t just some fancy add-ons. They genuinely boost conversion rates and help establish trust with users, all while keeping everyone safe. Plus, they really want to keep customers coming back for more! Want to learn more about this topic? Check it out here! You’ll find some great insights!

How 7Block Labs Works Its Magic

Discovery + Threat Model (48 hours) → We put together a solid plan to tackle any issues that matches up with your business timelines.
Targeted Engineering Sprints:
We're starting things off by diving into gas and compiler tuning. We’ll show you some pretty clear before-and-after results that really highlight the difference! We carefully go through a detailed hook risk review and make sure everything's patched up, all in line with Uniswap's guidelines. Check it out here.
When it comes to MEV, we're really zeroing in on routing and getting onboard with ERC‑4337. If you want to get into the nitty-gritty, just click here. You'll find all the details you need! We’re also diving into cross-chain hardening with tools like CCIP and LayerZero V2. On top of that, we’re setting up some rate-limited lanes to boost security even more. Learn more here. We get ready for audits at the same time, launch bounty programs, and work on estimating the costs for Layer 2 decentralized apps.

If your team needs a little extra support, we’re totally here for you! Whether you want us to take the reins or collaborate with your engineers, we've got your back either way. Just let us know how we can help! Take a look at what we have in store for you!

Full-stack protocol and L2 engineering: Check out our awesome custom blockchain development services and our web3 development services. We're here to help you navigate the exciting world of blockchain!
Building DeFi Products: Don't worry, we’ve got your back! Check out our awesome DeFi development services, or take a look at our DEX development and smart contract development offerings. We're here to help you make your DeFi dreams a reality!
Security and audit readiness: Rest easy knowing you’ve got our security audit services backing you up.
Interop and bridges: Check out our awesome cross-chain solutions and dive into our bridge development services. We’ve got some great stuff lined up for you!
dApp and onboarding: If you're looking to dive into the world of dApps, take a look at our dApp development services to kick things off! We've got everything you need to get started.

Bottom line
It’s not just about rewriting things; it’s really about knowing when to pull the right strings. Think of it like using gas-aware compilation, keeping your hooks in check, managing private order flow, and using rate-limited cross-chain lanes--all at just the right moment. In just four weeks, we rolled out this project, and wow, it came with some seriously competitive fees and strong defenses. Plus, the metrics we hit really made a difference in our go-to-market strategy!

CTA for DeFi

Are you all set to jump into the world of DeFi? Don’t miss out--reserve your spot for our DeFi 4-Week Rescue Sprint Call!

References (selected)

So, here’s the scoop on Dencun and EIP-4844: they’re getting activated soon! We’re looking at a blob retention period of roughly 18 days. Mark your calendars for March 13, 2024--execution will kick off at epoch 269568. Exciting stuff ahead! (ethereum.org). So, EIP-1153 is introducing some new transient storage opcodes, which is pretty exciting! You’ll be able to use them in Solidity by diving into inline assembly--just think of tstore and tload when you’re working with it. (eips.ethereum.org). So, there’s this new thing called EIP-5656 that rolls out MCOPY. We’re diving into how it affects things, especially with Solidity 0 in the mix. 8. We've been diving into the 25's optimizer and the MCOPY adoption, and let me tell you, there are some really exciting gains happening through the IR pipeline! (eips.ethereum.org). Hey, just a heads up--EIP-6780 is making some changes to how SELFDESTRUCT works after Dencun. Basically, it means you won’t be able to redeploy by just wiping stuff out anymore. (eips.ethereum.org). Hey, you should take a look at the Flashbots Protect and MEV-Share docs! They’ve got some really cool info on ERC-4337 bundlers and EntryPoint. Plus, they're diving into CCIP rate limits and sharing some insights on risk management. Oh, and don’t miss out on the LayerZero V2 DVN thresholding details--they're pretty interesting too! (docs.flashbots.net). Chainalysis just dropped some new trendlines for 2024 and 2025 that really dive into theft profiles. They even touch on what went down with the Bybit incident, giving it some solid context. (chainalysis.com).

Just a heads up! Whenever we talk about things like fees, blobs, or theft trends in the industry, we always make sure to back it up with links to the original specs or reputable research. We want you to have reliable info to chew on! When it comes to the performance metrics for the project, we get that info directly from our engineering logs and the client dashboards.