Read every older project's site again, once, and re-judge its leads - #86
Merged
Merged
Conversation
…file nobody touched Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
… pages Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
profile_reseedjob, queued once per project at boot (2 min apart) for projects made before 2026-09-20. It reads the site again with the multi-page reader. A profile nobody touched (profileVersion == 1) is replaced whole; an edited one keeps every fact and only has empty exclusions / not-buyers filled. A changed profile queues arescore; dropped platform searches queue a new discovery.flat()no longer lets a link cut open by the 12k homepage limit swallow the text after it (live bug from Judge leads against who can use the product; read each page for who cannot #80: valid sources were dropped on long homepages).Measured (1,463 labelled prod posts, 79 projects)
Same within run-to-run noise (~14 flips). The gain is truthfulness: Opus fact-check of every limit against the site found 1 false / 10 trivial of 73 new exclusions, against 7 false / 61 trivial of 141 before.
Cost: about $0.004 per project to re-read plus about $0.01 to rescore, under $2 for all of prod, spread over about 4 hours.
Tests
New
tests/profileReseed.test.ts; site-reading and prompt tests extended.tests/jobs.test.ts"seeds a scan" times out locally against the shared dev DB (passes in 7s with a longer timeout), as before.🤖 Generated with Claude Code