메뉴
Use when repeated work is governed by an explicit metric, active baseline, candidate comparison, promotion gate, leaderboard feedback, or submission iteration; do not use for a one-off research check that its current owner can execute directly.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use when a user message needs Chinese-first intent inference, low-confidence collaborative clarification, routing, workflow weight control, skill choice, review/testing/planning/debugging/evaluation routing, proactive multi-thread collaboration or conflict coordination, continuation handling, execution permission inference, or protection from process over-escalation.
Use when active development or explicit refactoring has concrete code-structure risk: a source file is accumulating independently changing responsibilities, a class/component/function is hard to navigate or test, repeated changes touch distant sections, or the user asks to split a giant file; do not use for line count alone, generated/vendor/migration/schema/fixture files, or cohesive modules that remain easier to maintain together.
Use when the user explicitly enables an automated same-project multi-thread development loop that keeps peer Codex tasks progressing, recovers conflicts or stale coordination, integrates verified work into local main, and closes cleanup without repeated gates; do not use before opt-in, for one-off coordination, remote publication without permission, or subagent creation.
Use when existing independent Codex App threads work in the same project and need peer advice, collaboration proposals, shared progress awareness, ownership arbitration, conflict pause/resume, discussion, merge coordination, status exchange, or handoff through native thread tools without creating subagents.
Use when a verified or integrated software change has concrete temporary artifacts, generated noise, stale claims, merged local branches, or disposable worktrees to clean, or when the user explicitly asks to remove development residue or old branches.
Use when the user or project rules explicitly request project-local memory/dashboard initialization or updates, or when an existing `.docs/project-memory/` is needed for cross-session recovery, durable decisions, blockers, parallel-agent coordination, or formal handoff; do not auto-initialize or sync it for ordinary development.
| name | ccdawn-score-loop |
| description | Use when repeated work is governed by an explicit metric, active baseline, candidate comparison, promotion gate, leaderboard feedback, or submission iteration; do not use for a one-off research check that its current owner can execute directly. |
| license | MIT |
围绕一个明确 metric 连续比较 baseline 与 candidate，并用可复现证据决定 PROMOTE / REJECT / HOLD。这是单条量化 lane owner，不负责整个研究方向或竞赛生命周期。
ccdawn-ai-research-loop、返回 ccdawn-competition-research-lifecycle、ccdawn-bug-review、ccdawn-brt 或 BLOCKED。ccdawn-brt 或更具体 owner，复合任务不吞其他 owner。下一步建议: <一个具体动作>。只有主要未知量是“候选能否持续改善明确 metric，并需要晋升/淘汰循环”时进入。AI Research 中可在同一上下文完成的一次低成本实验由研究 owner 直接执行，不为一次比较加载本 skill。
分数下降、candidate reject 和 online neutral/worse 是实验结果，不是 TDD RED。确定性的 metric/parser/schema/seed/shape/NaN/打包 bug 临时路由 ccdawn-bug-review，修复后返回原 lane。
QUICK：已有 baseline、命令和 gate，直接比较一个 candidate；不创建 profile、ledger、worker 或独立 artifact。STANDARD：反复候选、跨会话或需要晋升历史；维护最小记录。FULL：leaderboard/online feedback、提交包、昂贵运行或并行 lane；使用项目 profile 和持久 artifact。缺少正式 profile 不阻塞 QUICK；只要 metric、baseline、命令、范围和 gate 足以解释结果即可。STANDARD/FULL 才持久化这些字段，优先读取项目已有事实源，不创建平行 ledger。
smallestDecisiveEvaluation 和 kill condition。PROMOTE / REJECT / HOLD / BLOCKED。一个 lane 只改变一个可归因机制。Smoke/proxy 用于淘汰和排序，不替代目标评估；online score 是稀疏外部证据，不覆盖干净本地协议。没有新的因果变量、诊断信号或校准价值时停止重复尝试。
默认不创建 worker、worktree 或 lane matrix。只有候选写入面独立、资源允许且并行收益明显时才隔离；worker 只返回 diff、命令、指标和 artifact，不能自行晋升共享 baseline。
由 AI Research 发起时只回传：Candidate / Hypothesis outcome / Metric evidence / Mechanism evidence / Caveat / Reusable lesson / Pivot signal，研究方向仍由研究 owner 决定。由 Competition Lifecycle 发起时回传 gate 与提交/榜单影响。
Gate: PROMOTE / REJECT / HOLD / BLOCKED
- Baseline / Candidate:
- Metric delta 与关键副作用:
- 机制判断与证据:
- 记录/Artifact: <仅需要时>
下一步建议: <一个具体动作>