Issues

Report agent wall separately from grading_ms in run.md and the eval board
UiPath/coder_eval#212 · opened on Oct 2, 2026
Support voice/video/native image inputs
UiPath/coder_eval#202 · opened on Sep 28, 2026
Add Github Copilot CLI harness
UiPath/coder_eval#199 · opened on Sep 24, 2026
Invitation to share coder_eval on GithubStarMate
UiPath/coder_eval#185 · opened on Sep 17, 2026
no way to seal an agent's network without also sealing its model access, so no-internet tasks fall back to success criteria
UiPath/coder_eval#168 · opened on Sep 14, 2026
Allow dashboard customization
UiPath/coder_eval#167 · opened on Sep 11, 2026
4
Antigravity: gemini's task.json doesn't have message_id.
UiPath/coder_eval#163 · opened on Sep 10, 2026
codex agent: error notifications (incl. willRetry) are dropped by _CodexTurnState.dispatch — an 18-minute provider retry loop left no trace in task.json/task.log
UiPath/coder_eval#151 · opened on Sep 2, 2026
1
Add auto-annotation mechanism!
UiPath/coder_eval#135 · opened on Aug 24, 2026
Proposal: an EvalPort/OpenEval adapter for Coder Eval's EvaluationResult / TaskResult
UiPath/coder_eval#134 · opened on Aug 23, 2026
Support report-only criteria without discarding their declared weight
UiPath/coder_eval#101 · opened on Aug 10, 2026
Claude Bedrock model_used records SDK alias instead of effective route model
UiPath/coder_eval#99 · opened on Aug 10, 2026
Turn-count deltas are only valid within a model — reports must pin same-model baselines
UiPath/coder_eval#98 · opened on Aug 10, 2026
Create a task/experiment init command
UiPath/coder_eval#30 · opened on Jul 20, 2026
1