If you’re in a telegram group where they post alpha and you’re always missing it. Reply here or send a dm if you will need help setting it up I have an open source bot that can help you trade those calls in 10ms. You can also whitelist the specific accounts you trade their call

You can inspect the implementation, report a concrete issue, discuss the actual execution path with maintainers, and see fixes translated into code and regression tests. With v2.2 on the horizon, I’m glad to have contributed to that process.

That combination of implementation, integration, and verification is important for a protocol’s release process. This is why I value contributing to Percolator.

The wider v2.2 work is moving alongside these individual fixes. Recent repository activity includes engine integration work in PR #300⁠ and restoration of the Kani verification build for the v2.2 RC23 branch in PR #301⁠.

Issue #127⁠ github.com/dcccrypto/percola… documented settlement behavior that could destroy accumulated positive PnL beyond the loss being settled. The issue records a fix in d448e1c9 and an extensive regression suite covering the reported behavior.

The resolution review was closed as not exploitable through the wrapper, with additional hardening deferred. These aren’t the only examples of the engineering work around Percolator.

The maintainers traced the actual on-chain call paths and concluded that the public wrapper rejects the repeated-resolution path. That distinction matters too. Not every suspicious code path turns out to be an exploitable vulnerability.

The third report examined repeated market resolution. Issue #125 github.com/dcccrypto/percola… The concern was that resolving an already-resolved market could preserve payout snapshot and ledger state from an earlier resolution.

The project documented regression tests for the side-flip cases, including an equal-size flip, and reported that the relevant tests failed when the fix was reverted. The fix was also ported across both v2.2 engine lines.

The project reproduced the behavior and documented a fix that explicitly classifies sign flips and opens from flat as risk-increasing. The response went beyond changing a condition.

But changing sides is not equivalent to simply reducing exposure. Why did that matter? The classification feeds into risk-increasing trade checks, including stale-price and oracle-target-lag restrictions, as well as initial-margin source-lien creation.

The second report concerned position flips. Issue #126⁠ github.com/dcccrypto/percola… The engine’s risk classifier could treat a long-to-short or short-to-long flip as non-risk-increasing when the absolute position size stayed the same or decreased.

The fix introduced try_clear_bankruptcy_hlock_if_healthy() and connected it to liquidation, refresh, and settlement paths. The maintainers documented checks against v2.2 release candidates and deployed devnet v2 code before closing the issue.

The first was a bankruptcy-state recovery issue. #124⁠ github.com/dcccrypto/percola… The bankruptcy_hlock_active flag could remain enabled even after the underlying loss state had been settled. A market could recover economically while certain protocol operations remained locked.

Percolator is approaching v2.2, and I’ve been following the engineering work closely. My recent reports covered three different areas: • Risk classification during position flips • Recovery from settled bankruptcy states • Payout accounting during market resolution

One thing I appreciate about building in open-source DeFi is what happens when contributors and maintainers actually work together. I recently reported several edge cases in Percolator’s risk engine and settlement logic. A thread 🧵

I hope it does

A lot of fixes coming in v2.2 GitHub notification been going brahh