issues Search Results · language:Dune language:Python language:HTML language:Java language:HTML linked:pr language:Java
Filter by
4.5M results
Problem
The baseline gate (compare) fails CI on any raw score drop beyond a fixed tolerance. With non-deterministic agents and
small trial counts, small raw drops are often noise, and the gate produces ...
enhancement
arthurzengg
- [ ] The project is using the blacksmith.sh GitHub Runners instead of the standard ones
enhancement
good first issue
rijnhardtkotze
Environment
Hermes Desktop connected to a remote gateway/dashboard — desktop app on machine A, backend runtime on machine B (e.g. a
Linux mini-PC backend with the desktop app on a laptop over a private ...
Justlrnal4
[21] Improve create wizard UX with guided sections, colors, output choices, and final summary
!-- jules_eval template_id: v5_dynamic template_version: 5.1 scope_type: cli_ux_guarded eval_enabled: true ...
jules
wildersa
There is no test-coverage measurement or floor, so regressions in coverage go unnoticed.
Goals
- Measure test coverage.
- Enforce a minimum threshold ( = 85%).
- Include the coverage check in CI. ...
ci
Context
ci-self-hosted-runner has validated the python-ci runner against central-mcp-gateway. The next useful validation is a
second Python consumer: this repository.
Goal
Add a manual, opt-in workflow ...
vinicius-ssantos
概要
Codex 使用量の表示で、採用している rate_limits がいつのデータなのか分かるようにします。 あわせて、Codex セッションログ内の最新 rate_limits
イベント選択を強化し、古い表示に見える原因を診断しやすくします。
背景
AI Usage Tray の Codex provider は、公式 API ではなく ~/.codex/sessions のセッションログを読みます。 ...
enhancement
Mir2569
Make the static HTML report more useful for triage.
Goals
- Filter to only failed cases.
- Per-grader aggregation view (passed/total + rate).
- Improve transcript expand/collapse behavior.
- ...
enhancement
Goal
Reduce repeated valid-agent mask filtering in PredictiveMPPIAdapter._sequence_rollout() without changing rollout scoring
behavior.
Context
robot_sf/planner/predictive_mppi.py filters predicted ...
agent
evidence:smoke
performance
resource:local
state:ready
ll7
The agent-eval compare command exists but is not wired into CI. CI should fail when quality regresses against a
committed baseline, and always upload the run s reports.
Goals
- Compare current eval ...
ci
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.