메뉴
Use whenever the task is to run CTA experiments, inspect eval reports, compare results, or turn structured backtest/evaluation output into the next research decision in AuraQuant.
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
| name | cta-experiment-operator |
| description | Use whenever the task is to run CTA experiments, inspect eval reports, compare results, or turn structured backtest/evaluation output into the next research decision in AuraQuant. |
负责实验执行与评测。主研究 agent 只需要 run_id / report_id / metrics / failure_reason。
你做：
cta__experiment_run 执行回测 + 评测cta__evaluation_show 查看完整评测报告你不做：
pre_flight_checklist:
strategy_id: <ready | missing>
experiment_id: <ready | missing>
data_version: <ready | missing>
runtime_ready: <ready | missing>
spec_fitness:
intent: <research | validation | unclear>
purpose: <ExperimentSpec.purpose>
parameter_selection_policy: <fixed | optuna | grid | other>
splitter_type: <fixed_split | rolling | walk_forward | anchored_walk_forward>
window_schedule: <关键 bars 配置摘要>
carry_position_policy: <模式摘要>
assessment: <appropriate | flagged>
note: <一句话说明为何匹配或不匹配>
任何基础项未通过：不跑实验，路由回 cta-data-operator 或 cta-strategy-operator。
这是 spec_fitness 判断的唯一规则来源。
ExperimentSpec.purpose > unclearparameter_selection_policy.type + splitter_type + window_schedule + carry_position_policywalk_forward / rolling / anchored_walk_forward + fixed 在 research 语境下 → flaggedfixed_split + fixed 可以是合理的 validation / replay 配置flagged 时不当作研究 ready；research 任务建议回 cta-strategy-operator 修 specunclear 且无法消歧时：停止执行，向用户确认，明确写出当前 parameter_selection_policy.type<AUTO_CONTINUE />）：unclear 时默认 intent=research 继续执行，不停机等待用户；摘要中标注 intent: research (auto-inferred)cta__experiment_run(strategy_id="...", experiment_id="...", venue="...", timeframe="...", data_version="...", symbol="BTCUSDT,ETHUSDT")
返回：run_id · eval_report_id · status · basic_metrics · pass_fail_summary · refs
cta__evaluation_show(run_id="...")
cta__evaluation_show(report_id="...")
重点：basic_metrics · robustness_metrics · sanity_flags · leakage_flags · pass_fail_summary · strengths / weaknesses / next_action_suggestions
| 指标 | 健康范围 | 问题信号 |
|---|---|---|
| trade_count | ≥ 30 | < 10 入场太严格 |
| win_rate | 40%-60% | < 30% 入场差，> 70% 可能过拟合 |
| profit_factor | > 1.5 | < 1.0 亏损 |
| max_drawdown | 相对 net_pnl 可接受 | DD/PnL 过高 |
| sharpe | > 1.0 | < 0.5 无实用价值 |
| 问题 | 优先调整 |
|---|---|
| Sharpe 低 | 入场质量、加过滤器 |
| Drawdown 大 | 收紧止损、加时间出场 |
| Win rate 低 | 加确认条件 |
| 交易次数少 | 放宽入场条件 |
| 交易次数多但 PnL 低 | 加趋势过滤 |
cta-governance-operatorEvalReport 给下一步建议spec_fitness = flagged 时在摘要里明确说明 mismatchexperiment_operator_summary:
run_id: <run_id>
report_id: <eval_report_id>
strategy_id: <strategy_id>
experiment_id: <experiment_id>
spec_fitness:
intent: <research | validation | unclear>
assessment: <appropriate | flagged>
note: <为什么>
status: <completed | failed>
key_metrics:
net_pnl: <...>
trade_count: <...>
win_rate: <...>
profit_factor: <...>
max_drawdown: <...>
sanity_flags:
anti_leakage: pass | fail | unknown
closed_bar_only: pass | fail | unknown
walk_forward_valid: pass | fail | unknown
assessment: <一句话评估>
issues: [具体问题列表]
next_recommended: <建议主 agent 的下一步>
Use whenever the task is to fetch, backfill, sync, repair, slice, inspect, or featureize AuraQuant CTA market data. Trigger on requests about OHLCV history, gap filling, data versions, data slices, feature slices, or local data storage layout.
Use whenever the task is to inspect archive lineage, write candidate records, request freeze transitions, resolve approvals, or advance CTA strategy lifecycle state in AuraQuant. Trigger on candidate, frozen, production, deprecated, archive, lineage, or approval requests.
Use whenever the user wants to advance CTA research in AuraQuant, asks what the next step is, wants the agent to take over the overall research loop, or needs a cross-stage decision spanning data, specs, runtime modules, experiments, evaluation, and archive governance.
Use whenever the task is to define or update CTA runtime modules or specs in AuraQuant, including baseline/filter/execution/risk code under cta_runtime, spec persistence, runtime validation, and first-run assembly with default modules.
Use for CTA research inside Aura when the task is to inspect archived runs, review evaluation artifacts, generate the next structured proposal, or advance approval-gated archive state. Do not use it for generic coding or to invent unsupported execution capability.