Repository navigation
bench 268, audit 7: held rows stay out of the pool, effective sample size, single-chain floor - #2507
Merged
Merged
Conversation
… (Banana Gun was published on 10 Solana swaps weighted two thirds plus its fee-free Ethereum buys); the pooled entry needs 25 effective swaps (Kish); a single-chain product publishes at the row floor so the main view agrees with its tab; attempts never below the successes seen; copy Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01L4rjxemTvgRwrCG6D2iK5D
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Seventh loop of the bench audit (verdict NOT READY, report in .baudit/). What this fixes:
Held rows re-entered the pool (blocker). When no chain row of a product was firm, the pooled entry fell back on every row, held ones included: Banana Gun was published and ranked #6 at 364 bps on 10 Solana swaps carrying two thirds of the weight plus 208 fee-free one-sided Ethereum buys (effective sample 22). The pooled entry now waits when no row is published on its own, and a pooled entry also needs 25 effective swaps (Kish's (Σw)² / Σw², published as
n_eff).BasedBot read unresponsive on the main view at 36 priced while its Robinhood Chain tab published the same swaps at the chain floor. A single-chain product's pooled entry now follows its row's floor (25), so the main view agrees with the tab.
Attempts on Robinhood Chain undercounted a quarter of the flow. The block sample only counts transactions sent to the router; swaps through a smart account or a relayer emit the router's event without naming it as
to. Attempts are never below the successes the log feed saw plus the estimated failures; the hub tooltip says the EVM counts are estimated from sampled blocks.Copy. Publication rules in one bullet (50 pooled, 25 per chain or single-chain product, 25 effective, the one-side and main-chain holds); EVM rows at 1,000 draws a day; FOMO's legs "2 % or its $0.10 minimum"; Banana Gun's Ethereum row described as waiting for its sells and fee instead of asserting an off-chain fee; pump.fun fees 1 to 1.4 %; BasedBot in the intro; the Pool description and tooltip say a stable-to-gas-coin hop's own cost is excluded on the EVM rows (its rate values the native leg): valuing that hop at its pool's pre-trade state is the next harness change.
Deploy: harness, then worker rebuild + materialize reset.
🤖 Generated with Claude Code
https://claude.ai/code/session_01L4rjxemTvgRwrCG6D2iK5D