Preserve measurement targets when liquidity publication is rejected - #1042
Merged
Conversation
The failed source candidate replaced 582 active measurement targets with 21 before liquidity persistence rejected it. Prepare target descriptors during scoring and publish them only after all guards and the full liquidity generation succeed, preserving the previous catalog on rejected or failed candidates. Plan: agents/2026-09-05-night-watch/README.md Tasks: T1, T2
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A failed DefiLlama source cycle published a reduced measurement target catalog (582 → 21 targets; 468 → 11 pools) before the liquidity publication rejected the candidate. Subsequent collectors successfully measured only that reduced inventory, allowing broader exit evidence to age and lowering safety grades.
Prepare target descriptors during scoring, then publish active/shadow targets only after source and coverage guards pass and the full liquidity generation persists. Rejected candidates and failed writes preserve the previous catalog. Existing cadence, timestamps, quote identity validation, and scoring thresholds are unchanged.
Validation: 50 focused tests, all 441 cron/scheduled files (5,110 tests), Worker typecheck, lint, cron parity/connection checks, documentation checks, and full generated-artifact convergence passed. Independent review found no blocking issues. Regression coverage includes failed sources, coverage/value rejection, failed persistence, successful publication ordering, and shadow publication.
Runtime acceptance: verify the next full DEX publication rebuilds broad targets, then verify the next measurement collector consumes them and compare all 13 previously A-rated assets. Deployment alone does not close the incident.