Do you really want to be the last chain without network-native security?

No lock-in. Because once the network integrates, they can run the software forever, and the dApps can write assertions forever, all without our help. A security solution that isn't neutral isn't worth it.

Audits are essential. Monitoring is crucial. Prevention is the missing piece.

- verifiable security & financial ratings for dapps & assets (soon)

And I have heard... ...there is a demo.

You can now hook up the @phylaxsystems docs to your favorite AI tools, like Cursor or Claude. Makes writing assertions simple. See below:

Another hack, I've looked at recently is Vicuna Finance from March this year. They lost $700K through LP token oracle manipulation on the Sonic chain. The attack exploited a fundamental pricing flaw that could have been prevented with a simple assertion. LP tokens were priced using a basic sum formula (price_token1 * amount_token1 + price_token0 * amount_token0) instead of fair pricing that accounts for the pool's constant product formula. Attack sequence: - Large swap from token0 to token1 artificially inflated LP token oracle price - Deposit overvalued LP tokens as collateral - Borrow maximum assets against inflated collateral value - Reverse swap deflated LP price, leaving the protocol with bad debt Price manipulation is a pattern we see over and over again, and it's a pattern that assertions protect well against. In this example, we're experimenting with a new cheatcode that will allow for inspection of calls in the callstack, which is ideal for detecting intra tx price manipulations. We check that a "swap" call cannot result in the price deviating more than 5% from the baseline at any time during tx execution. It's a simple, but powerful way to protect against, for example, flash loan attacks.

The best example of catching the hack rather than monitoring the attack vector. Watch closely.

Assertions are enforced by the network itself. Not by a monitoring service that might go down. Not by a dashboard that might miss something. Not by an AI model that might guess wrong. By the same consensus mechanism that validates every transaction. The same validators that secure billions in value. The same network that has never been successfully attacked. When your assertion says "this invariant must hold," the network enforces it. Period. If a transaction would violate your security property, it doesn't get a warning. It doesn't get flagged for review. It doesn't execute. The network itself *is* your security layer.

super excited and honored to contribute to @OptimismGov, especially when it comes to doing it with frens