gavindiaz/EA_with_Python
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add safe-stop diagnostics to H1 entry logs

Browse Source
This commit is contained in:
Hiroaki86
2026-06-04 19:56:13 +09:00
parent 33cfc2597c
commit 13eaa74309
9 changed files with 1139 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python_for_ea/src/ea_py/io/debug_logs.py
+28
View File
@@ -103,3 +103,31 @@ def append_debug_entry(
file.write("---- REASON START ----\n")
file.write((reason_text or "").strip() + "\n")
file.write("---- REASON END ----\n\n")
def append_debug_entry_stop(
*,
path: Path,
stage: str,
reason: str,
trend_state: int,
selected_strategies: Sequence[int] = (),
candidate_id: str = "",
current_price: float | None = None,
imbalance_summary: str = "",
) -> None:
"""H1候補生成が安全側停止値へ倒れた理由を追記する。"""
with path.open("a", encoding="utf-8") as file:
file.write("=" * 60 + "\n")
file.write(f"DEBUG TIME : {now_str()}\n")
file.write("RESULT : SAFE_STOP\n")
file.write(f"STAGE : {stage}\n")
file.write(f"REASON : {reason}\n")
file.write(f"MARKET_STATE(H4) : {trend_state} ({MARKET_STATE_LABELS.get(trend_state, 'UNKNOWN')})\n")
file.write(f"SELECTED_STRATEGY : {','.join(str(x) for x in selected_strategies) or '-'}\n")
file.write(f"CANDIDATE_ID : {candidate_id or '-'}\n")
if current_price is not None:
file.write(f"CURRENT PRICE : {current_price:.2f}\n")
if imbalance_summary:
file.write(f"IMBALANCE(H1) : {imbalance_summary}\n")
file.write("=" * 60 + "\n\n")
python_for_ea/src/ea_py/io/debug_logs.py.bak_codex_20260601_debug_entry
+105
View File
@@ -0,0 +1,105 @@
"""デバッグ理由ログをMT5 Files配下へ追記する。"""
from __future__ import annotations
from datetime import datetime
from pathlib import Path
from typing import Sequence
from ea_py.constants import MARKET_STATE_LABELS
def now_str() -> str:
"""デバッグログ用の現在時刻文字列を返す。"""
return datetime.now().strftime("%Y-%m-%d %H:%M:%S")
def append_debug_trend(
*,
path: Path,
model: str,
reasoning_effort: str,
max_output_tokens: int,
api_diagnostics: str,
current_price: float,
numeric_summary: str,
direction_numeric: str,
market_state: int,
classification_reason: str,
reason_text: str,
) -> None:
"""H4トレンド判定の数値要約、方向出力、分類理由を追記する。"""
with path.open("a", encoding="utf-8") as file:
file.write("=" * 60 + "\n")
file.write(f"DEBUG TIME : {now_str()}\n")
file.write(f"MODEL : {model}\n")
file.write(f"REASONING : {reasoning_effort}\n")
file.write(f"MAX TOKENS : {max_output_tokens}\n")
file.write(f"CURRENT PRICE : {current_price:.2f}\n")
file.write(f"MARKET STATE : {market_state} ({MARKET_STATE_LABELS.get(market_state, 'UNKNOWN')})\n")
file.write("=" * 60 + "\n\n")
file.write("---- API DIAGNOSTICS START ----\n")
file.write((api_diagnostics or "").strip() + "\n")
file.write("---- API DIAGNOSTICS END ----\n\n")
file.write("---- NUMERIC SUMMARY START ----\n")
file.write((numeric_summary or "").strip() + "\n")
file.write("---- NUMERIC SUMMARY END ----\n\n")
file.write("---- DIRECTION NUMERIC (0/1/2) START ----\n")
file.write((direction_numeric or "").strip() + "\n")
file.write("---- DIRECTION NUMERIC (0/1/2) END ----\n\n")
file.write("---- MARKET STATE CLASSIFICATION START ----\n")
file.write((classification_reason or "").strip() + "\n")
file.write("---- MARKET STATE CLASSIFICATION END ----\n\n")
file.write("---- REASON START ----\n")
file.write((reason_text or "").strip() + "\n")
file.write("---- REASON END ----\n\n")
def append_debug_entry(
*,
path: Path,
model: str,
reasoning_effort: str,
max_output_tokens: int,
api_diagnostics: str,
timeframe: str,
current_price: float,
trend_state: int,
selected_strategies: Sequence[int],
imbalance_summary: str,
numeric_summary: str,
numeric_lines: str,
post_filter_summary: str,
sanitized_numeric_list: Sequence[int | float] | None,
reason_text: str,
) -> None:
"""H1エントリー候補生成の数値行と理由を追記する。"""
with path.open("a", encoding="utf-8") as file:
file.write("=" * 60 + "\n")
file.write(f"DEBUG TIME : {now_str()}\n")
file.write(f"MODEL : {model}\n")
file.write(f"REASONING : {reasoning_effort}\n")
file.write(f"MAX TOKENS : {max_output_tokens}\n")
file.write(f"TIMEFRAME : {timeframe}\n")
file.write(f"MARKET_STATE(H4) : {trend_state} ({MARKET_STATE_LABELS.get(trend_state, 'UNKNOWN')})\n")
file.write(f"SELECTED_STRATEGY : {','.join(str(x) for x in selected_strategies)}\n")
file.write(f"IMBALANCE(H1) : {imbalance_summary}\n")
file.write(f"CURRENT PRICE : {current_price:.2f}\n")
file.write("=" * 60 + "\n\n")
file.write("---- API DIAGNOSTICS START ----\n")
file.write((api_diagnostics or "").strip() + "\n")
file.write("---- API DIAGNOSTICS END ----\n\n")
file.write("---- NUMERIC SUMMARY START ----\n")
file.write((numeric_summary or "").strip() + "\n")
file.write("---- NUMERIC SUMMARY END ----\n\n")
file.write("---- GPT NUMERIC LINES START ----\n")
file.write((numeric_lines or "").strip() + "\n")
file.write("---- GPT NUMERIC LINES END ----\n\n")
file.write("---- POST FILTER START ----\n")
file.write((post_filter_summary or "").strip() + "\n")
if sanitized_numeric_list is not None:
file.write("sanitized_numeric_list=" + ",".join(str(x) for x in sanitized_numeric_list) + "\n")
file.write("---- POST FILTER END ----\n\n")
file.write("---- REASON START ----\n")
file.write((reason_text or "").strip() + "\n")
file.write("---- REASON END ----\n\n")
python_for_ea/src/ea_py/pipelines/entry_pipeline.py
+112 -14
View File
@@ -26,7 +26,7 @@ from ea_py.constants import (
TECHNICAL_ERROR_STOP,
USE_IMBALANCE_FILTER,
)
from ea_py.io.debug_logs import append_debug_entry
from ea_py.io.debug_logs import append_debug_entry, append_debug_entry_stop
from ea_py.io.mt5_files import read_int_file, write_results_then_done
from ea_py.io.ohlc_csv import read_ohlc_csv
from ea_py.market.volatility import summarize_ohlc
@@ -48,7 +48,7 @@ from ea_py.market.imbalance import (
format_imbalance_summary,
)
from ea_py.openai_client import call_responses_api, create_openai_client
from ea_py.paths import build_entry_paths
from ea_py.paths import EntryPaths, build_entry_paths
from ea_py.prompts.entry_prompt import (
build_imbalance_guidance,
build_caution_block,
@@ -67,6 +67,41 @@ logger = logging.getLogger(__name__)
TREND_STOP_STRATEGIES = frozenset({1, 3})
def exception_summary(error: Exception) -> str:
"""秘密情報を追加せず、例外型とメッセージを1行へ丸める。"""
return " ".join(f"{type(error).__name__}: {error}".splitlines()).strip()
def write_safe_stop(
*,
paths: EntryPaths,
stage: str,
reason: str,
trend_state: int,
selected_strategies: Sequence[int] = (),
candidate_id: str = "",
current_price: float | None = None,
imbalance_summary: str = "",
) -> None:
"""停止値を出力し、ライブ診断用の理由ログも可能な限り残す。"""
if DEBUG_PRINT:
try:
append_debug_entry_stop(
path=paths.debug_reason,
stage=stage,
reason=reason,
trend_state=trend_state,
selected_strategies=selected_strategies,
candidate_id=candidate_id,
current_price=current_price,
imbalance_summary=imbalance_summary,
)
except Exception:
logger.exception("debug_entry.txt safe-stop write error")
write_entry_output(stop_numeric_list())
def write_entry_output(numeric_list: NumericList, zone_text: str | None = None) -> None:
"""target_prices.txtとtarget_zones.txtを書き、完了後にprocess_done_entry.txtを作る。"""
paths = build_entry_paths()
@@ -124,19 +159,37 @@ def run_pipeline() -> None:
trend_state,
MARKET_STATE_LABELS.get(trend_state, "UNKNOWN"),
)
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="market_state",
reason="H4 market_state blocks new entries.",
trend_state=trend_state,
selected_strategies=selected_strategies,
)
return
try:
ohlc_all = read_ohlc_csv(paths.input_csv)
except Exception:
except Exception as error:
logger.exception("CSV 読み込みエラー")
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="ohlc_csv",
reason=exception_summary(error),
trend_state=trend_state,
selected_strategies=selected_strategies,
)
return
if len(ohlc_all) < CANDLE_LONG:
logger.error("データ本数不足: len(ohlc)=%s (need >= %s)", len(ohlc_all), CANDLE_LONG)
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="ohlc_bars",
reason=f"Insufficient H1 bars: actual={len(ohlc_all)} required={CANDLE_LONG}.",
trend_state=trend_state,
selected_strategies=selected_strategies,
)
return
ohlc_short = ohlc_all[-CANDLE_SHORT:]
@@ -164,7 +217,16 @@ def run_pipeline() -> None:
MARKET_STATE_LABELS.get(trend_state, "UNKNOWN"),
imbalance_summary,
)
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="h1_imbalance",
reason="H1 imbalance conflicts with the allowed H4 strategies.",
trend_state=trend_state,
selected_strategies=selected_strategies,
candidate_id=candidate_id,
current_price=current_price,
imbalance_summary=imbalance_summary,
)
return
try:
@@ -186,9 +248,18 @@ def run_pipeline() -> None:
png_file_to_data_url(paths.tmp_short_chart),
png_file_to_data_url(paths.tmp_long_chart),
]
except Exception:
except Exception as error:
logger.exception("画像生成エラー")
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="chart_render",
reason=exception_summary(error),
trend_state=trend_state,
selected_strategies=selected_strategies,
candidate_id=candidate_id,
current_price=current_price,
imbalance_summary=imbalance_summary,
)
return
numeric_summary = build_numeric_summary(
@@ -219,9 +290,18 @@ def run_pipeline() -> None:
try:
config = load_runtime_config(debug_print=DEBUG_PRINT)
except RuntimeError:
except RuntimeError as error:
logger.exception("OpenAI 設定エラー")
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="openai_config",
reason=exception_summary(error),
trend_state=trend_state,
selected_strategies=selected_strategies,
candidate_id=candidate_id,
current_price=current_price,
imbalance_summary=imbalance_summary,
)
return
system_content = build_system_content_block_debug() if config.debug_print else build_system_content_block()
@@ -248,11 +328,29 @@ def run_pipeline() -> None:
api_diagnostics = gpt_result.diagnostics.to_log_text()
if not gpt_result.diagnostics.is_completed():
logger.error("OpenAI response incomplete or failed: %s", api_diagnostics)
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="openai_response",
reason=api_diagnostics,
trend_state=trend_state,
selected_strategies=selected_strategies,
candidate_id=candidate_id,
current_price=current_price,
imbalance_summary=imbalance_summary,
)
return
except Exception:
except Exception as error:
logger.exception("OpenAI APIエラー")
write_entry_output(stop_numeric_list())
write_safe_stop(
paths=paths,
stage="openai_api",
reason=exception_summary(error),
trend_state=trend_state,
selected_strategies=selected_strategies,
candidate_id=candidate_id,
current_price=current_price,
imbalance_summary=imbalance_summary,
)
return
if config.debug_print:
python_for_ea/src/ea_py/pipelines/entry_pipeline.py.bak_codex_20260601_debug_entry
+305
View File
@@ -0,0 +1,305 @@
"""H1 OHLCとH4 market_stateからtarget_pricesを生成するパイプライン。"""
from __future__ import annotations
import logging
from collections.abc import Sequence
from ea_py.charting.candlestick import ohlc_to_candlestick_png_file, png_file_to_data_url
from ea_py.config import load_runtime_config
from ea_py.constants import (
CANDLE_LONG,
CANDLE_SHORT,
DEBUG_PRINT,
ENTRY_MAX_DISTANCE_LIMIT_EATR_MULTIPLIER,
ENTRY_MAX_DISTANCE_MIN_PRICE,
ENTRY_MAX_DISTANCE_STOP_EATR_MULTIPLIER,
ENTRY_MAX_OUTPUT_TOKENS,
ENTRY_TIMEFRAME,
IMBALANCE_AVG_BODY_PERIOD,
IMBALANCE_MIN_AVG_BODY_SIZE,
IMBALANCE_SENSITIVITY,
INSTRUMENT,
MARKET_STATE_LABELS,
MIN_ENTRY_REWARD_RISK_RATIO,
MT_ENCODING,
TECHNICAL_ERROR_STOP,
USE_IMBALANCE_FILTER,
)
from ea_py.io.debug_logs import append_debug_entry
from ea_py.io.mt5_files import read_int_file, write_results_then_done
from ea_py.io.ohlc_csv import read_ohlc_csv
from ea_py.market.volatility import summarize_ohlc
from ea_py.market.target_prices import (
stop_numeric_list,
strategies_by_trend,
)
from ea_py.market.target_zones import (
build_candidate_id,
format_target_zones,
parse_json_to_entry_zones,
sanitize_entry_zones,
stop_entry_zones,
zones_to_numeric_list,
)
from ea_py.market.imbalance import (
adjust_strategies_for_imbalance,
detect_imbalance_signal,
format_imbalance_summary,
)
from ea_py.openai_client import call_responses_api, create_openai_client
from ea_py.paths import build_entry_paths
from ea_py.prompts.entry_prompt import (
build_imbalance_guidance,
build_caution_block,
build_common_rules_block,
build_entry_response_text_format,
build_header,
build_market_state_guidance,
build_numeric_summary,
build_system_content_block,
build_system_content_block_debug,
)
from ea_py.types import NumericList
logger = logging.getLogger(__name__)
TREND_STOP_STRATEGIES = frozenset({1, 3})
def write_entry_output(numeric_list: NumericList, zone_text: str | None = None) -> None:
"""target_prices.txtとtarget_zones.txtを書き、完了後にprocess_done_entry.txtを作る。"""
paths = build_entry_paths()
content = "".join(f"{number}\n" for number in numeric_list)
zone_content = zone_text if zone_text is not None else format_target_zones(stop_entry_zones(), "0")
write_results_then_done(
result_files=[
(paths.output_prices, content, MT_ENCODING),
(paths.output_zones, zone_content, MT_ENCODING),
],
done_path=paths.done_entry,
)
def build_entry_distance_limits(h1_eatr: float, selected_strategies: Sequence[int]) -> dict[int, float]:
"""H1 EATRから戦略別のエントリー許容距離を作る。"""
limits: dict[int, float] = {}
for strategy in selected_strategies:
multiplier = (
ENTRY_MAX_DISTANCE_STOP_EATR_MULTIPLIER
if strategy in TREND_STOP_STRATEGIES
else ENTRY_MAX_DISTANCE_LIMIT_EATR_MULTIPLIER
)
limits[strategy] = max(h1_eatr * multiplier, ENTRY_MAX_DISTANCE_MIN_PRICE)
return limits
def run_pipeline() -> None:
"""H1 OHLCとH4 market_stateからMT5向けtarget_pricesを生成する。
実行契約:
- `trend_state.txt` のH4 `market_state` を読み、許可戦略だけを選ぶ。
- `ohlc_H1.csv` から短期・中期チャート画像と数値要約を作る。
- OpenAIには選択済み戦略の候補価格だけを依頼し、返却行を13値形式へ展開する。
- GPT出力は `sanitize_numeric_list` で検証し、`target_prices.txt` を書いた後に
`process_done_entry.txt` を作成する。
安全側フォールバック:
H4が技術エラー停止値、CSV読込失敗、H1データ不足、画像生成失敗、OpenAI設定/API失敗、
GPT出力不正、または全候補が価格条件違反の場合は13値すべてを停止値にする。
Pythonは候補価格を作るだけで、注文送信やポジション操作は行わない。
"""
paths = build_entry_paths()
trend_state = read_int_file(
paths.trend_state,
encoding=MT_ENCODING,
allowed_values=MARKET_STATE_LABELS.keys(),
default=TECHNICAL_ERROR_STOP,
)
selected_strategies = strategies_by_trend(trend_state)
if trend_state == TECHNICAL_ERROR_STOP or not selected_strategies:
logger.info(
"market_state=%s (%s). New entries stopped.",
trend_state,
MARKET_STATE_LABELS.get(trend_state, "UNKNOWN"),
)
write_entry_output(stop_numeric_list())
return
try:
ohlc_all = read_ohlc_csv(paths.input_csv)
except Exception:
logger.exception("CSV 読み込みエラー")
write_entry_output(stop_numeric_list())
return
if len(ohlc_all) < CANDLE_LONG:
logger.error("データ本数不足: len(ohlc)=%s (need >= %s)", len(ohlc_all), CANDLE_LONG)
write_entry_output(stop_numeric_list())
return
ohlc_short = ohlc_all[-CANDLE_SHORT:]
ohlc_long = ohlc_all[-CANDLE_LONG:]
current_price = float(ohlc_short[-1]["Close"])
candidate_id = build_candidate_id(str(ohlc_short[-1].get("DateTime", "")))
imbalance_analysis = detect_imbalance_signal(
ohlc_all,
avg_body_period=IMBALANCE_AVG_BODY_PERIOD,
sensitivity=IMBALANCE_SENSITIVITY,
min_avg_body_size=IMBALANCE_MIN_AVG_BODY_SIZE,
)
imbalance_summary = format_imbalance_summary(imbalance_analysis)
selected_strategies = adjust_strategies_for_imbalance(
selected_strategies,
trend_state=trend_state,
analysis=imbalance_analysis,
use_filter=USE_IMBALANCE_FILTER,
)
if not selected_strategies:
logger.info(
"H1 imbalance conflicts with market_state=%s (%s). New entries stopped. %s",
trend_state,
MARKET_STATE_LABELS.get(trend_state, "UNKNOWN"),
imbalance_summary,
)
write_entry_output(stop_numeric_list())
return
try:
ohlc_to_candlestick_png_file(
ohlc_data=ohlc_short,
save_path=paths.tmp_short_chart,
instrument=INSTRUMENT,
timeframe=ENTRY_TIMEFRAME,
dark=True,
)
ohlc_to_candlestick_png_file(
ohlc_data=ohlc_long,
save_path=paths.tmp_long_chart,
instrument=INSTRUMENT,
timeframe=ENTRY_TIMEFRAME,
dark=True,
)
images_data_urls = [
png_file_to_data_url(paths.tmp_short_chart),
png_file_to_data_url(paths.tmp_long_chart),
]
except Exception:
logger.exception("画像生成エラー")
write_entry_output(stop_numeric_list())
return
numeric_summary = build_numeric_summary(
current_price=current_price,
ohlc_short=ohlc_short,
ohlc_long=ohlc_long,
)
short_summary = summarize_ohlc(ohlc_short)
h1_eatr = float(short_summary.get("eatr", 0.0))
max_entry_distance = build_entry_distance_limits(h1_eatr, selected_strategies)
post_filter_summary = (
"entry_distance_guard="
f"max_distance_by_strategy={max_entry_distance}, "
f"h1_eatr={h1_eatr:.2f}, "
f"stop_multiplier={ENTRY_MAX_DISTANCE_STOP_EATR_MULTIPLIER:.2f}, "
f"limit_multiplier={ENTRY_MAX_DISTANCE_LIMIT_EATR_MULTIPLIER:.2f}, "
f"floor={ENTRY_MAX_DISTANCE_MIN_PRICE:.2f}, "
f"min_reward_risk={MIN_ENTRY_REWARD_RISK_RATIO:.2f}"
)
header = build_header(
current_price=current_price,
numeric_summary=numeric_summary,
trend_state=trend_state,
)
market_state_guidance = build_market_state_guidance(trend_state)
imbalance_guidance = build_imbalance_guidance(imbalance_analysis)
caution = build_caution_block()
try:
config = load_runtime_config(debug_print=DEBUG_PRINT)
except RuntimeError:
logger.exception("OpenAI 設定エラー")
write_entry_output(stop_numeric_list())
return
system_content = build_system_content_block_debug() if config.debug_print else build_system_content_block()
common_rules = build_common_rules_block(selected_strategies, max_entry_distance)
response_text_format = build_entry_response_text_format(selected_strategies)
max_tokens = ENTRY_MAX_OUTPUT_TOKENS
user_text = "\n\n".join([header, market_state_guidance, imbalance_guidance, common_rules, caution]).strip()
api_diagnostics = ""
try:
client = create_openai_client(config.api_key)
gpt_result = call_responses_api(
client=client,
model=config.model,
reasoning_effort=config.reasoning_effort,
system_content=system_content,
user_text=user_text,
image_data_urls=images_data_urls,
max_output_tokens=max_tokens,
response_text_format=response_text_format,
)
gpt_reply = gpt_result.text
api_diagnostics = gpt_result.diagnostics.to_log_text()
if not gpt_result.diagnostics.is_completed():
logger.error("OpenAI response incomplete or failed: %s", api_diagnostics)
write_entry_output(stop_numeric_list())
return
except Exception:
logger.exception("OpenAI APIエラー")
write_entry_output(stop_numeric_list())
return
if config.debug_print:
logger.info(
"---- TREND ---- %s selected=%s\n---- GPT REPLY START ----\n%s\n---- GPT REPLY END ----",
trend_state,
selected_strategies,
gpt_reply,
)
entry_zones = parse_json_to_entry_zones(gpt_reply)
entry_zones = sanitize_entry_zones(
entry_zones,
selected_strategies,
current_price,
max_entry_distance=max_entry_distance,
min_reward_risk_ratio=MIN_ENTRY_REWARD_RISK_RATIO,
)
numeric_list = zones_to_numeric_list(entry_zones)
zone_text = format_target_zones(entry_zones, candidate_id)
if config.debug_print:
try:
append_debug_entry(
path=paths.debug_reason,
model=config.model,
reasoning_effort=config.reasoning_effort,
max_output_tokens=max_tokens,
api_diagnostics=api_diagnostics,
timeframe=ENTRY_TIMEFRAME,
current_price=current_price,
trend_state=trend_state,
selected_strategies=selected_strategies,
imbalance_summary=imbalance_summary,
numeric_summary=numeric_summary,
numeric_lines=gpt_reply,
post_filter_summary=post_filter_summary,
sanitized_numeric_list=numeric_list,
reason_text=f"STRUCTURED_JSON\n{gpt_reply}\n\nTARGET_ZONES\n{zone_text}".strip(),
)
except Exception:
logger.exception("debug_entry.txt write error")
write_entry_output(numeric_list, zone_text)
def main() -> None:
"""ログ設定を行ってH1エントリー候補生成パイプラインを起動する。"""
logging.basicConfig(level=logging.INFO, format="%(levelname)s:%(name)s:%(message)s")
run_pipeline()
python_for_ea/tests/unit/test_debug_logs.py
+31
View File
@@ -0,0 +1,31 @@
"""Debug log output tests."""
from __future__ import annotations
from pathlib import Path
from ea_py.io.debug_logs import append_debug_entry_stop
def test_append_debug_entry_stop_records_safe_stop_reason(tmp_path: Path) -> None:
"""Safe-stop diagnostics distinguish a guarded stop from a normal GPT skip."""
path = tmp_path / "debug_entry.txt"
append_debug_entry_stop(
path=path,
stage="openai_api",
reason="RateLimitError: retry later",
trend_state=2,
selected_strategies=[1, 2],
candidate_id="202606012100",
current_price=4535.17,
imbalance_summary="signal=NONE",
)
actual = path.read_text(encoding="utf-8")
assert "RESULT : SAFE_STOP" in actual
assert "STAGE : openai_api" in actual
assert "REASON : RateLimitError: retry later" in actual
assert "MARKET_STATE(H4) : 2 (LOW_VOL_UP)" in actual
assert "SELECTED_STRATEGY : 1,2" in actual
assert "CANDIDATE_ID : 202606012100" in actual