issues Search Results · language:Dune language:Python language:JavaScript language:TypeScript language:TypeScript linked:pr
Filter by
7.8M results
Parent bounty: #743
This is an independent reissue of #4467 because the original issue is limited to its creator and asks non-authors to
create another issue with the same contents.
apps/api/src/app.js ...
Problem
Graders today mostly check the final outcome. Many agentic tasks need credit for reaching the right intermediate states
in the right order (e.g. identity must be verified before a refund is processed). ...
enhancement
Overview
Build a public, guest-accessible /bracket page that shows the FIFA World Cup 2026 knockout bracket (Round of 32 → Final)
from real match data, plus a logged-in Compare mode that grades the user ...
enhancement
Problem
A pass/fail flag and a free-text grader reason do not tell you how trials fail across a run. We want a structured
failure taxonomy with first-bad-step localization and cross-run aggregation.
...
enhancement
Parent bounty: #33
Bug
apps/web/src/app/page.tsx exports no metadata object. Next.js App Router uses exported metadata for title and meta name=
description tags. Without it, the homepage falls back ...
Problem
A single suite-level pass@k number hides how reliability scales with attempts and which tasks are flaky. We want the
full curve and a flakiness view.
Acceptance criteria
- Report pass@1..N ...
enhancement

Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip! Restrict your search to the title by using the in:title qualifier.