Repository Issue Activity (beta)

#209 [Bug] codex_exec backend: rollouts skipped (calls=0) + bwrap sandbox blocks file reads despite danger-full-access

#56 Feat: Add Windows support for opencode plugin

#155 Proposal: AgenticReplayBackend — replay mined tasks with a real agentic CLI in a throwaway worktree, gate on the diff

#192 关于选择 LiveMathematicianBench 作为评测数据集的疑问

#194 `merge_batch_size: 1` causes an infinite loop in `_hierarchical_merge`

#197 CodexCliBackend passes the prompt via argv: the Windows .CMD shim truncates it at the first newline, so every codex rollout silently scores 0

#174 Gate compares slice averages only: an aggregate gain can hide a per-task regression, and improvements are never attributed to the rule that caused them

#175 section_present rejects numbered or annotated headings, and the optimizer 'fixes' it by forbidding the skill's own heading format

#145 关于目前skill进化的疑惑。目前好像很多算法都是仅针对SKILL.md做优化吗？

#195 wrong issue

#158 Several reported scores are incompatible with the released test split sizes

#67 Add a verifier-discipline stress test for reward-hacking skill edits

#154 Mining is limited to programmatic success checks — intent-level improvements get dropped or Goodharted into shallow proxy rules

#157 LiveMath scores imply 125 test items, but the released split has 124

#132 Help wanted: integrate SkillOpt with Superpowers for measurable agent efficiency

#164 consolidate._split leaks held-out test-split tasks into train and val when a batch is entirely test-split

#94 Strengthen verifier-discipline test to assert held-out scores and gate_action (follow-up to #87)

#119 Can lcoal LLM be integrated to the harness?

#156 Proposal: community rule library with local empirical gating (privacy-preserving federated skill transfer)

#113 A Confusion Regarding s_cur and s_best

#111 Explore SkillOpt for improving Cursor agent skills from our session data

#120 SkillOpt-Sleep: target existing hand-written skills, not just the dedicated managed skill

#90 Clarification request: Qwen enable_thinking setting used for reported ALFWorld results

#97 SkillOpt-Sleep: Support harvest from Vscode/ Visual studio copilot sessions

#38 Subject: Clarification on Slow Update vs. Meta Skill design