Simmer Calibration Report

核心功能是对Simmer JSONL交易日志（实盘、模拟均可），按策略、时段、价格区间等维度拆解胜率、每笔交易EV，明确交易优势与短板。适合交易者复盘优化策略，输入可配置日志路径、回溯天数等参数，输出结构化校准报告，支持简洁/详细模式，新手易上手，仅需安装依赖、设置APIKey即可运行。

djdyll

热度 53.04

最近更新 2026-05-11 17:31:38

分类：数据依赖：代码执行 / 读写文件 / 第三方 API应用场景：数据处理 / 报告撰写 / 财务

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add simmer-calibration-report

Metadata

name: simmer-calibration-report
displayName: Simmer Calibration Report
description: Run a calibration report on any Simmer trade journal. Win rate and EV broken down by strategy, time of day, price band, and market type. Know exactly where your edge lives — and where it doesn't.
version: 1.0.4
authors: [{"name":"DjDyll"}]
difficulty: beginner

📊 Simmer Calibration Report

You think you know where your edge comes from. This skill tells you if you're right.

This is a template. Point it at any Simmer JSONL trade journal — live, sim, or exported from anywhere — and get a full calibration breakdown. No strategy-specific logic baked in. Your journal, your numbers.

Setup

Install dependencies:
```
pip install simmer-sdk
```
Set your API key:
```
export SIMMER_API_KEY=your_key_here
```
Run it:
```
python calibration_report.py --live
```

That's it. If your journal exists at the default path, it just works.

Configuration

Parameter	Env Var	Default	Description
`journal_path`	`CALIB_JOURNAL_PATH`	`""`	Path to trade journal JSONL. Empty = auto-detect.
`min_trades`	`CALIB_MIN_TRADES`	`10`	Minimum trades to show a segment. Below this, the numbers lie.
`lookback_days`	`CALIB_LOOKBACK_DAYS`	`30`	How far back to analyze.
`include_unresolved`	`CALIB_INCLUDE_UNRESOLVED`	`false`	Include unresolved trades. EV becomes less meaningful.

Set config inline:

python calibration_report.py --set lookback_days=60
python calibration_report.py --set min_trades=5

Quick Commands

# Sim journal (dry run data)
python calibration_report.py

# Live journal
python calibration_report.py --live

# Custom journal path
CALIB_JOURNAL_PATH=/path/to/journal.jsonl python calibration_report.py

# Show config
python calibration_report.py --config

# Summary only (good for cron)
python calibration_report.py --live --quiet

Example Output

📊 Calibration Report — Last 30 days
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Total: 287 trades | 68.3% WR | +$0.038 EV/trade
Journal: data/live/trade_journal.jsonl

BY STRATEGY
  btc_momentum         n=142  WR=72.5%  EV=+0.047  ██████████████
  eth_midcandle        n=67   WR=61.2%  EV=+0.021  ██████
  btc_midcandle_scalp  n=49   WR=65.3%  EV=+0.031  ████████
  eth_btc_lag          n=29   WR=58.6%  EV=+0.015  ████

BY HOUR (UTC)
  00-06                n=44   WR=75.0%  EV=+0.055  ██████████████████
  06-12                n=89   WR=70.8%  EV=+0.044  ████████████████
  12-18                n=97   WR=65.0%  EV=+0.031  ██████████
  18-24                n=57   WR=64.9%  EV=+0.028  █████████

BY ENTRY PRICE
  0.20-0.35            n=38   WR=73.7%  EV=+0.052  ██████████████████
  0.35-0.50            n=71   WR=70.4%  EV=+0.041  █████████████
  0.50-0.65            n=89   WR=66.3%  EV=+0.033  ██████████
  0.65-0.80            n=61   WR=63.9%  EV=+0.024  ████████

🏆 Best segment: btc_momentum (EV=+0.047, WR=72.5%)

Troubleshooting

Problem	Fix
"No journal found"	Set `CALIB_JOURNAL_PATH` explicitly, or ensure `data/live/trade_journal.jsonl` exists in your workspace.
0 resolved trades	Trades haven't resolved yet. Set `CALIB_INCLUDE_UNRESOLVED=true` to see pending trades (EV will be approximate).
Segment not showing	Below `min_trades` threshold. Lower it: `python calibration_report.py --set min_trades=5`
"SIMMER_API_KEY not set"	`export SIMMER_API_KEY=your_key_here`
"simmer-sdk not installed"	`pip install simmer-sdk`