issues Search Results · language:Edge language:Python language:PHP language:TypeScript language:HTML language:JavaScript
Filter by
68.2M results
What to build
Implement the mechanics for validation benchmark fixtures and fixture-candidate queueing.
Ship a versioned authoritative seed fixture set, language-tagged fixture structure, fixture loading/selection ...
ready-for-agent
Heratiki
What to build
Define the human review and promotion policy that governs when fixture candidates move into the authoritative validate
benchmark set.
Establish ownership, rubric, required evidence, and ...
needs-triage
증상
모바일에서 특가 상세 LP의 하단 sticky CTA(StickyMobileCTA, 상담 신청 )가 사이트 전역 하단 네비(mobile-bottom-nav)에 가려 탭이 가로채입니다. 두 fixed 바가 화면 하단에
겹쳐(약 2단 적층) sticky CTA가 실질적으로 클릭 불가.
원인
- 전역 src/components/layout/mobile-bottom-nav.tsx ...
thxforall
What to build
Add a unified MCP tool for explicit feedback on generator and validator performance.
Feedback should update the corresponding model scores with stronger manual weighting than automatic ...
ready-for-agent
What to build
Wire synchronous validation into task execution with a precomputed trio retry ladder: self-validation pre-filter,
optional external validation, and retry escalation from to to without re-routing. ...
ready-for-agent
What to build
Add a dedicated validator-selection path () using confidence-blended validation scores and threshold gating.
Selector should prefer local/free tiers before paid, exclude the active generation ...
ready-for-agent
What to build
Introduce validation-scoring primitives and configuration gates needed for validator selection and retry policy.
Add validate as a benchmark task category, add scores.validate capability ...
ready-for-agent
What to build
Deliver a standalone module with a narrow interface that validates generated output against task requirements using a
YES/NO-first-line contract and graceful parser fallback.
Module should ...
ready-for-agent
What to build
Make free-tier trio ranking use the same -driven benchmark source as local-tier ranking, then extend ranking with
provider-diversity and live rate-limit factors so fallback candidates reduce ...
ready-for-agent
What to build
Update route_task and preemptive_route_task to return a ranked trio (good, better, best) in the response, including
benchmark metadata for each model and a fallback notice when fewer than ...
ready-for-agent
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.
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.