import marimo

__generated_with = "0.23.16"
app = marimo.App(width="medium", app_title="Vibe Coding 進階：讓測試當 AI 的眼睛（實驗場）")


@app.cell(hide_code=True)
def _(mo):
    mo.md(
        r"""
    # 🧪 Vibe Coding 進階：讓測試當 AI 的眼睛（實驗場）

    這是本課的**實驗場**。每個實驗都有下拉選單與滑桿可以操作，選完馬上重算。

    這裡的材料全部是**實測紀錄**：一個 2B 小模型（qwen3.5-2b，2026-09）寫了 12 道小程式題、
    每題重跑 8 次，每一版程式都丟進沙盒跑過 pytest。你看到的程式碼、測試輸出、通過與否，
    都是當時的原始紀錄；而第 3️⃣ 節的「測試設計師」會把這些程式**在你的瀏覽器裡重新跑一次**。
    """
    )
    return


@app.cell
def _():
    import marimo as mo
    return (mo,)


@app.cell
def _():
    # 共用套件集中在這格 import（其他格一律從參數拿，不重複 import）
    import base64
    import builtins
    import contextlib
    import difflib
    import html
    import io
    import itertools
    import json
    import zlib

    import matplotlib.pyplot as plt
    import numpy as np
    return base64, builtins, contextlib, difflib, html, io, itertools, json, np, plt, zlib


@app.cell
def _(base64, json, zlib):
    # ── 實測資料（spike_genai_vibecoding.py／_risks.py 的輸出，由 _pack.py 壓縮注入；不要手改）──
    VIBE_B64 = "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"
    RISK_B64 = "eNrtXXtz08ba/yr7uv/AvLGxHEhoZnhnKKRt5tA2B+jp6dRMqtibREWWhCRDDMNMwi030gQm4ZYUCjScBBquBdKQhA9TS3b+6ld4n92VbUmWZCfxCU7rwCS2pN199nl+z3VX0rlQCut8qO1cKCUnsRhqC506g6XmyL5wrDvUFEryOoZjsWisJRz9OBzbC8c0PqWIWAu1NTeFdJxSsMrraRUui0Za4YisdCnk8/6mkHKyt0tR5ZSiQx/m8LX8o8Hc1DzqzOh9soTO6bx28vwfAxfyT8bMH+fzNx6hZmQMjJmzA+uDC8bkcO7OJaQICjLmbmbfvflz5aox8Ti/smQMLdPDgqTpvCgi6BKuZBcZk+O5+RdwaXZpIH+f/IYO6ddxGDz77l12ddq4/Cb77oZ5Y8hYvAmjkxnhhIr1EqnGsyesZYFUY2jVWJlAWjqV4lXhLN6l4359N6Ho2oox8QR9pWDpYAciBB/s7EDm6Cjr3rx2C4Y1hydJdxNz+YUXMGBxFrn5Meg192CFUHGesivU9t250ElBSgIVKZ7MUMV8CkgkvIJj+acDQBxq709gEZmPp/5cGY70i1r/nysjRckA85tCKn8GLpeBLiXTL8YlhZeSvBaX+kU1CVcqfOIk30uECONJfArbLobTuF/QdDipq2ncFDqNVU2QJbikOcJF9pGBKCMycORggUWi0K3CIaTLCEhO7jmjCjq2CI1FuSgiZO6BXynUIxD4NIVEXtO70ooo80mGsb3haEs4th94USSK0R1IUjTS4iCpUz6D1Z60iAC8PAIOphMEnxrqkVV2jJd4MaMJWhPShRRGGlYFDF9gKLic12EgIeFJIFUCrtVOoMVQP/JiQF7MQd4Ri1GUGnwai8B3VSN8A0ypfEJnNPYAGNEXQkKVNblHtxi5S0/tRkTgSFMIl7U+jHV/fu4j/OT2hs6fIARvGVZcA1YNWNUeVrEGrBqw2giscrdX0OfHj3ciEji8uGD54KPtx44T7+vlB1V8Ko1hxnEprYog++a41KfrSr8PtgqXB3Kqea+LVRa0KGWEXZ+nU7ykRXwks49OtSQZi7DAIfdHoo4R6VAFKJ8R9D6k9xFGhzW+B6OELEk4oUNjpMiyKEi9TZT58E3TGX5Ssooj/tDZZyewwDA/8gA5+yOcg77jfRhJgBPUiyUSJQoF9iREAUu638j7w81cDVHANVDQQEHJyzRQ8NdDASRcqPPwpyj/8D7kYFbuNboKOVZ27YqXQ1CSPYqYTnVjFYKNZE9MSIHc4WMGvvgAodQkcOIcF+GckukkrRBPCWRuVOeB9UlIYOFrirGC75bTOsJ8og8l+ni1CcKShM5LvSJm0gGR+UqnxSWd4oQCCAUqW10IOojOqLwC7h3xqpyGMXUQGfQFGb2SokTQLwleUGUAFDSz6IJoQBZpTAAwgyF0OlX4CnPuOII6CClwoeYtZAijuHDUEaUUxOBHfEuE+7iMeAWiprDC9I6MX9CFBK/w3YB0uQfCEFHQdQr+FFZ76QcIWRSFfiLTgZBG0ohU4AgTmF+0wrSipXbY5Lyw2ZmBpjGKUSAJqxFN6N/x+GSTqhAlcx9WvLEwF6Omx6FWdhn4mmaYM2QQrU4ew1gKr2pEuYAOGtKfpZLyYhjTh1oBK9Yweg2jt81Gzxi8k11aBEyaN97mF+56ueAfBNzNxyXgd48SlxJ9wFMNh62jjZONkwEnfawVvTbQUO2NuRzLIdYh+kZWkxo6hntTEIwy6/S1LoDmCN6aECUKHIvaFZiSaBu9hxc1bDvvmFjZdZvXKM5Do2AkSGt4SVTqllXQWQCzCrT7J0Du2t7nvHSksw3BH3QEfEia2N9OVU5gTQO7FpjYbJ73MRfvKREWWuuS85TA7eJrJd2omNxPP8kuXzfGpo2VCWP2Rv7+vGcuJ8q6mIlLEFTQj0J3XNIwxBaq5BfV0BYBXGgFLrhcrYRIxTmsyWk1gSGK0TGpfgqnMSuAnha0NC8KZ5lURFvVlJUCquRTaQ7BdW3Ou95A2hJXj6xZ+4/qiGxL3AoI8ZpdQx4r1J550YMFvkFPrAp9C5R70dx5yBuC2YJoveT+9+BuU93ifEtyjzX0/a+g75KQ6MMlka8v3MwvjhgTz3ODD/PLT4zhK+Tr4s387Xtetp6LIMh2zsalWASVlgubI0jsT4l+0ocGFfL1yF7HVMHPikm62nZWhkwvSYIVgcxSgxxKTmJVonlTH3Qoq4QhvnyPOeq6dblmaXEuIO2MNHuvWBIeMNqEBPr3F0eQUgwLSomonOoG5rEjMFJsD/mriXqhRg0zEDGJZ46rGJOavX8NOtpcAyhxXlBiGXSY7HZKQ0BFESUofDKpwnTqHlYu6gOju48j0Qip+TutXXu/jiVyDS1oEKFoOlnKVgsCJhqP6fIzEJoWsQ+2mkHj7ZTZeehflIlGYk6EdXSe3rsHfrWgFC8JSlp0WFmPsTmwcNEwt78G+Ig58KEIUkaQKEYEbr9UJTBYq0Dzum+fy+GYLwgt5tjK+r1X+dVfzfEH5vx946ebe4y3c8bltxGfFfNWV3xPqAxemHEadQFmLfRkqMyt9X2ijKRwRH2eRFlPzDxOIl6BixP0iJd14WJECs31iwCyL+/5NWN2zPj9N/PmnDl7L786Yg5P5ufeG5cvwkHP9MJtG5jVsGmbj3WoU52sH7vlwZzyvHALIuSCRFgu050oyQbGPDFWIwDFbACyqXwZfuo+QvDVtPoFU5AI783lFx4wR567c2n94uv1wbvG+AA48j9Xhpv/N4ZyD1Z8dkQXdhYyiVYpx3rf1dcIRbcHRZzTHvRiWcnsPGNAyQ6UjDs2tRj/GZYTcpLkcsdlWeyW+30GbA1zsb8PKFxOoleWe0Uc5hXBWpcNsw1OccmW/dc5QHymECi4WDTqQs1ntBt6Q8wh2h4dqb5M5own6qhwEgCe/H8emncnLfC8fpFdGsjdXsnNkbuTskuLufEfAUUAHy+3JKVT1JTYvJMPTOiVFbTXmUt9moZO6YqRWKgTUgnwqsrTAo2SpjVEQaoolJYdtdm9jqPQjUOFc0cwBcgEZC71LqA6wvJfCSoxf6vC9wsNu1IDu8L4uJGqhbWBYHrJWLwJoYTx6MIfA/T+1PGL2eWrudtr2aURY3zQmJn5Y2AcIgyWla7fem+OjWRXnrGGXq7D2lZCxMt2otE4qRiWBu+YKZ0NqBj67EYg98AG7WTk9pOAy7UiaacxcNRohHOWKg9Z2+y+BjVDR7DUq/ehwyBaT6lxYY4DwXlEoeUz/hDSq1iaUjJn5SSVqZpRdDkpp/BOrU/RqQRWHhxz9Nf6WLMrwjxE2/WqvNInJKpchN1Hb7doDW2zwG1ZQmFNQcmcUiGfwkWZnxLqcUWhQGZgAT9Wtl/7n0cRaVW43QNkQsJuABrdsypmSm6BrQQe+9dnTai981gT6vzyM7YtFZMdqeBc5LQOrsSLXo5uBHbTW2Jm1Sqenx/OPb1ZzPtIefDpHWNwyLw5lx8eyj+czU0swuegipLlbHd6Lal+4oDNLA9sXI6cd2WwwIZGPXBLpYQPjaaaQiXmo/KVCjsNla9W5XdGnccYvJNfe2nOXAGcAGzgQ3b5em5micQAEzONNeR6X0OuWn5cQ347Wn6xCvLbqav+9bPYV7dFs6cDxsQN5tyNwbns+5+McXKXZP7VkHnjdvDCcFIAB7bzvbo1kQoUxjxuKCzg0+7ZiUMlHp2QQvwq+Xy449BXX1DCfUhqdqVmO2QZp1rseDxYpIAiLSGcFPSwiHnfTekf4sESdQ9aF9/8Cw4flxUcNEx3LVo3wzKTyqhM8bToh2ivBRBT0lN0a7I/bdHaAKUscbABJiMnuxM71cZ8EAwXOOY3IiQrrkrY4U/oKjTDBB3yq8OfHPKuf5FdpMHVwdzUg9zUPSb29VHymEkIPIylV+w5lOj4N/88irLvbpm/rhWKYF7+plT7U/hMqRjog4SKJbj9Lh7DyIdI+Y3eEl8qwvlWRN0xBhC1sUUBWeoR1BTq5KEpSQozJD3U0GmBPAPgOIVXR+eXXjW8VsLz6L6dV3P8b2GEK8dI4ZNwStX0esHIzqoLF1hXdZqxJRnGvGQYl+pLyU+plZbgnMJLJhH4lD5Z1SF01xKqoNBd+X1Yxf/jswDY4nrKxbbMscoNXutPbq1PvibPTrl9wRh+53N7cW8aumD3llreuxvzkBOCk9XktLLXR5iFZkGusQX0w+mpvjn2WQdzjMewCldSZ/X1lx3/9n08VmxnBXhlvPMjdG+Ecy9oHUuomNySmVB5xXaPme8DYqKttYGC39YbW/TTqPtuqe67fXFkLfAQ88/7ArZONnLAKmFZL9tPT5w/UXg0PBOazTVQJ9YW+v777xmJEJmlFOIXS5Cgj8QlFQ9eQNZJ9pj4OFydxD3up8m3xSUEP2yu6IB18S5gQ9dJnDkQD3371ddHu2A6Xf9o/7br8/aj7fHQbtYGRKHIkoahlfU0xUQfD79kQi8tr0VgFryOd7HryQ993D/02qvoYfK4fxBotxwPNdmuwJpGAHvgu9Ix8nMuHlJlEcdDbSge0jKajlPQDj4nZEknO53piW/lNJgFDAFDHxYVeruwBgDReaAuDuIO6DMN8CnrsSceOlZgGK2m9siiKJ8h1oZuRBIk+uwtcvOxSABIGiZwGzpHzp6HEUsDnmAfi8yDqUtFHgLrZCGBte+iJyIWByIFMiSQN0DmDAl46DNq1CQBQpKi1+YtfNHB4OCJjaMgniNCStDbVVVWSy20YLjEQ/Qf/ex8GQEBvvF8iLxq4PJcwOsIWFvz1bx5aeLPlTvm4i/sbQfG4FB+ftG8MF94q8H6+Gvj7Utz8WHuwfL65fHc6lP6YoUpc/bx+uCUMfx8feBe7seh9duTxvCbYsd2AtmMIxakAa2yFunFwNrTu+KhrzrbvzzYUQA4YBuhj5Cxdjd3/bnx4GL+6ag5vcQqLgVqrAHob6EHSbLuGqGtJHOVF0A//sWLaUwZDAMC3ebIGtCNnEMj+4hFHWO/ddXRaUnvrIEPgd4dKqpdudZVp3lB2ledtjDQOxqecA1ge3XHAcjnSyd32ydYvW6Q63F/AituLNsYpqigpsB6KkAbWgBI+SdjuflxY+2qcWU8v/CI4CdUTsmXMnmwlW2odvqHPkxQQ7hsKLAaZLfs1L31q8/yj38Bc4DP+/cblz5C2dX3JMF7dim7NgZa2IO6uohf6upCBw4Aj7u6yINLurqAyawTwv4uQYLAB5HzMA3j9yfm8LWyd4gU3/aRXbti7T24vEBedzL9G+CM7EAY/dl8NW1cv5r9/RZ8oBwo2ve0SPov2YDSsLuLCsAuK+MBO7zbbb1A6xx2K1at3dIzNPq2Th0RyAOEDwsJ+H1QynhbqzYSNOxG4f8jfz0sF2O6y3LlXj8ynt6paLks2zdxEbhomzt1CrvIsG2I2UD2kExz9tdi01Lz/PspY+aurTmhE7FRc68nzbuzHo2YzbQ1KtrO9dkBQu6FeTCYzFSasyPm9HDBlIKVqc5Aenp9YhgJzp4PwYD529Pm1O+5C8vMUhvPHq3fmSkMUuemq0bBw6YCCEFDDJnFAKFpI/aS87SXpU9MHR1Ws83Z30ZtKzV6ooa9u4mHmK0FS1aSrYoTWDiNk+z1DCRoIs9XKYDNSXQle2qN01Mc6ByoyC68+zzprhpD+REyl34Fy577ZZmaVqpl1PZ0UV0ljchDwk+lhcRJ1K3KZ2A0uR/9AFmqhmRSDCEzEPmzGZSUeyPoOJEh/CdBH+mmGPTRAF+HKJw+GSmtKrKGtUiAMbWRYQkz0HJ6x33Nm4v7qskJKppKl4Gz7NaF5U1ZS8t4sT7Aovx4z/jPmMtyBllN8LZgqAkxtAsgwFicdLWsaDqd5hJsoTkynr+4uj44ll2d9bSdFRInZ4z3t8yfmM/IvpvLLo35uVXIAOJxaduTpr1ByqOV6dGWMqKKEQUJRR8sk+Xcty8hBQLvHZQhkLfATT4yRkA/LpqjM+S9brazkGRAFxDHkLuJaQzqCV5w6PQqltrcnCW5Fr3xAGjJDz3O379qzrw3J2a8/Lzt8pFN5FkEJ9rJcL/PT/0nQL4qxPBetI2FR7j6AR9y2Ucg9Nz0bfPnJeP9RTBQ5pvXNAzfXMRRoyTMKmkBjzcUK1gqarV2OHlLVpg4ct+8iWoL1R/jlxfmNKDrTmDyFJxd7KtpVeQHjTSvQV3E5TI3m2C4XGax5rJJl+nZ0t9lZtd+yr/6zbyxYhVrbBPNz19Yv/GCvWOT2TRbQYe41WKCkl2ayl26D+n4+v1xR65TKghNPAMPDFbOmPnZnrznB64WEEKyeEhIcpNXcjNLxC75WzyrYEO5CCgknwWNvtITIjgqyiaaNJaHoJDSX7oPCbI10MKyOXyDHKHxAmO9lTv/pRIfiH55vYR2LajsGdk2k8VtS92oKPkqC0cbTWlcNSJHZvMRYhkLC7yryXTKs5piokLPMMlRcUoYJ+n+4G5cEm0SdbMH8vWkJfrOpgjq0MlrJnTyzFyUgpxEsOc7Gnt/k9Dbp5N+dFlGogypD12ZI+0SpPZpOYJiHsQMvXcSVJbySGDiHfa8pbqAbcvLIbWqb28h72GvCa5sxD2rRc4gj5ngcuNbHkMyLAZnNoHV8+qitqrTne00gtnVn81bz1j11Hg+bM4tuty0Mbpgjowx3ScB/cRkdm2GBNjTb9lluZdrucsj7vDOFtvlXw3l50fhYO75fXN1CtJNY3RmW0M9N9f4/i5dPokl7QAXjdbSqFZh/hjYSIC34ZoOK5Szyg5LV4qJir0UiTPF5MRtHbPvRo1fFuC/+fyhsTZszK2CWEgBfeCa8eM74/Kvxux1c/qhuUi8eu7SG7D0xbP5l2+zyzPm899A5UAjjadvbOZ/QwbOHa+2/ncSUs/qTU1K3FuxXK7ws1K85uWYC5d5pLO5J8/W3901rl8tJqrsourTU/+VbywlFVmgxtGnk08OHmvv+vroEdYTeXGh1rZnD4wdscI/MHt7TnOuNPejgrOBLspqRFXGjTuwQtQpYl6zRSK+C+3k7p3gNfYmO049ktzdrtXbIn9JLBNQOq9lNuysprsr6Z+yF5bBbK1dh84YKth+7K9gP2q8O6SaUKCxS6Rq8Nao1ukGxcfbs0Swuarn1lYGNuxkKhfwPxBsy6I/TyB7XbaZIJaVZoqr3e4gFqLTgCAWpLC2QMLgx3fz8wNloayTxnJlCpyppV4V5knQ4RqmGh/A1UjBWFp64vz/A9QeRz8="
    V = json.loads(zlib.decompress(base64.b64decode(VIBE_B64)))
    RISK = json.loads(zlib.decompress(base64.b64decode(RISK_B64)))
    return RISK, V


@app.cell
def _(V):
    # 條件的顯示名稱與顏色（教學頁用同一組顏色）
    CONDS = [c for c in ["reroll", "bare", "full", "hint", "weak"] if c in V["tasks"][0]["b"]]
    COND_LABEL = {
        "reroll": "開新對話重抽",
        "bare": "只說「不對，再修」",
        "full": "貼回 pytest 錯誤原文",
        "hint": "錯誤原文＋你的一句診斷",
        "weak": "貼錯誤，但只有基本測試",
    }
    COND_EN = {"reroll": "re-roll (new chat)", "bare": "\"still wrong, fix it\"",
               "full": "paste pytest output", "hint": "output + human diagnosis",
               "weak": "paste output, basic tests only"}
    COND_COLOR = {"reroll": "#8172B2", "bare": "#9AA7AE", "full": "#55A868", "hint": "#4C72B0", "weak": "#C44E52"}
    PROB = {p["name"]: p for p in V["problems"]}
    RUNS = V["meta"]["runs"]
    MAXK = V["meta"]["max_attempts"]
    NPROB = len(V["problems"])

    def n_cases(name):
        return len(PROB[name]["basic"]) + len(PROB[name]["edge"])

    def att_at(rows, k):
        """第 k 次嘗試時手上的那一版（提早過關或停手，就一直是最後那版）。"""
        return rows[min(k, len(rows)) - 1]

    def is_pass(name, row):
        return row[1] + row[2] == n_cases(name)
    return (COND_COLOR, COND_EN, COND_LABEL, CONDS, MAXK, NPROB, PROB, RUNS, att_at, is_pass, n_cases)


@app.cell
def _(builtins, contextlib, io, json):
    # 瀏覽器版的「沙盒」：import 白名單、擋 input/open、吞掉 print。
    # 與 _spikes/spike_genai_vibecoding_pack.py 裡的同名函式邏輯相同——打包時已用它逐案例對照過 pytest，結論一致。
    ALLOWED = {"re", "math", "datetime", "collections", "decimal", "itertools", "functools", "string",
               "typing", "calendar", "fractions", "operator", "unicodedata", "bisect", "heapq", "enum", "numbers"}

    def safe_exec(code, name):
        def guarded_import(mod, globals=None, locals=None, fromlist=(), level=0):
            if mod == "pytest":  # 有的版本多 import 了測試框架卻沒用到：給一個空模組，行為與沙盒一致
                return type(builtins)("pytest")
            if mod.split(".")[0] not in ALLOWED and not mod.startswith("_"):  # _strptime 等標準庫內部模組放行
                raise ImportError(f"blocked import: {mod}")
            return __import__(mod, globals, locals, fromlist, level)

        def blocked(*a, **k):
            raise RuntimeError("blocked in sandbox")

        safe_builtins = dict(vars(builtins), __import__=guarded_import, input=blocked, open=blocked,
                             exit=blocked, quit=blocked)
        ns = {"__builtins__": safe_builtins, "__name__": "ai_solution"}
        try:
            with contextlib.redirect_stdout(io.StringIO()):
                exec(compile(code, "solution.py", "exec"), ns)  # noqa: S102 — 重跑錄下來的 AI 程式就是這格的目的
            return ns.get(name)
        except BaseException:  # noqa: BLE001 — 語法錯、import 被擋、模組層程式出錯，都算這份程式不能用
            return None

    def check_one(fn, args, expected):
        if fn is None:
            return False
        try:
            with contextlib.redirect_stdout(io.StringIO()):
                got = fn(*[json.loads(json.dumps(a)) for a in args])
        except Exception as e:  # noqa: BLE001 — 任何例外都要拿來跟「期望 raise 的型別」比對
            return isinstance(expected, dict) and type(e).__name__ == expected["raises"]
        if isinstance(expected, dict):
            return False
        return got == (tuple_fix(expected) if isinstance(expected, list) else expected)

    def tuple_fix(expected):
        """JSON 沒有 tuple：top_k_words 的期望值是 list[tuple]，還原回 tuple 才能與 pytest 的判定一致。"""
        return [tuple(x) if isinstance(x, list) and len(x) == 2 and isinstance(x[0], str) else x for x in expected]
    return check_one, safe_exec


@app.cell(hide_code=True)
def _(V, mo):
    mo.md(
        rf"""
    ## 1️⃣ 第一版沒過之後：五種下一步，誰救得回來？

    實驗設計（{V["meta"]["date"]} 實測）：

    - **需求只給一句話**，像 vibe coding 那樣——例如「寫一個 `round_half_up(x)`：把 x 四捨五入成整數」。
      細節（負數怎麼進位、錯誤輸入要不要 raise、輸出格式）**只寫在測試裡**，就像真實世界裡細節只存在你腦中。
    - 每題有兩層測試：**基本測試**（跟需求裡的例子同一類的快樂路徑）與**邊界測試**（需求沒講、但你心裡其實有答案的細節）。
    - 模型：`{V["meta"]["model"]}`（temperature {V["meta"]["temperature"]}），12 題 × 每題重跑 {V["meta"]["runs"]} 次。
      第 1 版五種條件**共用同一份**；沒全過，就用下列方式最多再試 3 次：

    | 下一步 | 做法 |
    |---|---|
    | 開新對話重抽 | 不給任何回饋，同一句需求重新生成；測試只負責判定哪一版過關 |
    | 只說「不對，再修」 | 同一段對話裡回一句「測試沒有全部通過，請修正」 |
    | 貼回 pytest 錯誤原文 | 同一段對話裡把失敗的測試輸出整段貼回去 |
    | 錯誤原文＋你的一句診斷 | 同上，再加一句「人看完失敗測試後會說的話」（只講哪裡錯、應該怎樣，不給程式碼） |
    | 貼錯誤，但只有基本測試 | 同上上，但 AI 只看得到基本測試——基本測試全綠就停手 |

    評分一律用**完整測試**（基本＋邊界）。下圖的線是 {V["meta"]["runs"]} 次重跑的平均、色帶是最少到最多——
    **你自己跑，數字會落在不同位置，看方向**。
    """
    )
    return


@app.cell
def _(mo):
    metric = mo.ui.dropdown(
        options=["全部測試都過的題數（滿分 12 題）", "通過的測試案例比例（全部 75 個案例）"],
        value="全部測試都過的題數（滿分 12 題）",
        label="看哪一種成績",
    )
    metric
    return (metric,)


@app.cell
def _(COND_COLOR, COND_EN, CONDS, MAXK, NPROB, RUNS, V, att_at, is_pass, metric, n_cases, np, plt):
    _by_solved = metric.value.startswith("全部")
    _total_cases = sum(n_cases(p["name"]) for p in V["problems"])
    curves = {}
    for _c in CONDS:
        _m = np.zeros((RUNS, MAXK))
        for _t in V["tasks"]:
            for _k in range(1, MAXK + 1):
                _row = att_at(_t["b"][_c], _k)
                _m[_t["r"], _k - 1] += is_pass(_t["p"], _row) if _by_solved else (_row[1] + _row[2]) / _total_cases
        curves[_c] = _m
    _fig, _ax = plt.subplots(figsize=(6.4, 4.6))
    _x = np.arange(1, MAXK + 1)
    _hi = 0
    for _j, _c in enumerate(CONDS):
        _m = curves[_c]
        _dx = (_j - (len(CONDS) - 1) / 2) * 0.035  # 平均值相同的線稍微錯開，才不會互相蓋住
        _ax.fill_between(_x, _m.min(0), _m.max(0), color=COND_COLOR[_c], alpha=0.12, linewidth=0)
        _ax.plot(_x + _dx, _m.mean(0), "-o", color=COND_COLOR[_c], label=COND_EN[_c], linewidth=2.2, markersize=5)
        _hi = max(_hi, _m.max())
    _ax.set_xticks(_x)
    _ax.set_xlabel("attempt # (1 = shared first attempt)")
    if _by_solved:
        _ax.set_ylabel(f"problems passing ALL tests (of {NPROB})")
        _ax.set_ylim(0, min(NPROB, max(6, _hi + 1)))
    else:
        _ax.set_ylabel("share of test cases passed")
        _ax.set_ylim(0, 1)
    _ax.set_title(f"{V['meta']['model']}, {RUNS} runs: mean line, min-max band", fontsize=10)
    _ax.grid(alpha=0.3)
    _ax.legend(fontsize=8.5, loc="upper center", bbox_to_anchor=(0.5, -0.16), ncol=2, frameon=False)
    _fig.tight_layout()
    _fig
    return (curves,)


@app.cell
def _(COND_LABEL, CONDS, MAXK, RUNS, V, att_at, is_pass, itertools, mo):
    _failed = [t for t in V["tasks"] if not is_pass(t["p"], t["b"][CONDS[0]][0])]
    _first_ok = len(V["tasks"]) - len(_failed)
    _rows = []
    for _c in CONDS:
        _rescued = sum(is_pass(t["p"], att_at(t["b"][_c], MAXK)) for t in _failed)
        _per_run = [sum(is_pass(t["p"], att_at(t["b"][_c], MAXK)) for t in V["tasks"] if t["r"] == r)
                    for r in range(RUNS)]
        _same = _rep = 0
        for _t in V["tasks"]:
            _r = _t["b"][_c]
            for _a, _b in itertools.pairwise(_r):
                _rep += 1
                _same += _a[0] == _b[0]
        _rows.append(
            f"| {COND_LABEL[_c]} | **{_rescued}** | {min(_per_run)}–{max(_per_run)} | "
            f"{_same}/{_rep}" + (f"（{_same / _rep:.0%}）" if _rep else "") + " |"
        )
    _table = "\n    ".join(_rows)
    mo.md(
        f"""
    第 1 版：{len(V["tasks"])} 個題次（12 題 × {RUNS} 次重跑）只有 **{_first_ok}** 個全過，剩下 **{len(_failed)}** 個交給各種下一步：

    | 下一步 | 救回幾個（共 {len(_failed)} 個） | 每輪 12 題最後全過（最少–最多） | 新版與上一版**一字不差** |
    |---|---|---|---|
    {_table}

    最後一欄最值得看：同一段對話裡，小模型很常把**上一版原封不動再交一次**——
    它「看得見」錯誤訊息，卻沒有拿它去改程式。（上圖可以切換成「通過的測試案例比例」，看部分進步。）
    """
    )
    return


@app.cell(hide_code=True)
def _(mo):
    mo.md(
        r"""
    ## 2️⃣ 逐版重播：AI 每一輪到底看到什麼、改了什麼

    挑一題、挑一次重跑、挑一種下一步，把每一版程式、當時的測試結果（每個圓點是一個測試案例，
    空心＝邊界測試）、AI 收到的回饋、以及**跟上一版的 diff** 攤開來看——
    這就是你用 AI 寫程式時該養成的習慣：**看 diff，不看它說了什麼**。
    """
    )
    return


@app.cell
def _(COND_LABEL, CONDS, V, mo):
    _names = [p["name"] for p in V["problems"]]
    trace_prob = mo.ui.dropdown(options=_names, value="round_half_up", label="題目")
    trace_cond = mo.ui.radio(
        options={COND_LABEL[c]: c for c in CONDS},
        value=COND_LABEL["full"], label="第一版沒過之後",
    )
    mo.hstack([trace_prob, trace_cond], justify="start", gap=2, wrap=True)
    return trace_cond, trace_prob


@app.cell
def _(RUNS, V, is_pass, mo, trace_prob):
    _opts = {}
    for _t in sorted(V["tasks"], key=lambda t: t["r"]):
        if _t["p"] != trace_prob.value:
            continue
        _first = _t["b"]["full"][0]
        _tag = "第 1 版就全過" if is_pass(_t["p"], _first) else "第 1 版沒過"
        _opts[f"第 {_t['r'] + 1} 次重跑（{_tag}）"] = _t["r"]
    _pref = [k for k, v in _opts.items() if v == 4 and "沒過" in k]  # 教學頁 hero 用的是第 5 次重跑
    _default = _pref[0] if _pref else next((k for k in _opts if "沒過" in k), next(iter(_opts)))
    trace_run = mo.ui.dropdown(options=_opts, value=_default, label=f"哪一次重跑（共 {RUNS} 次）")
    trace_run
    return (trace_run,)


@app.cell
def _(PROB, V, difflib, html, mo, trace_cond, trace_prob, trace_run):
    _p = PROB[trace_prob.value]
    _t = next(t for t in V["tasks"] if t["p"] == trace_prob.value and t["r"] == trace_run.value)
    _rows = _t["b"][trace_cond.value]
    _nb, _ne = len(_p["basic"]), len(_p["edge"])

    def _dots(row):
        # 每個圓點＝一個測試案例（依題目的測試順序）：實心＝基本、空心＝邊界；綠過紅沒過
        _b = V["bits"][row[0]] or [0] * (_nb + _ne)
        _s = "".join(
            f'<span title="basic {i + 1}" style="display:inline-block;width:11px;height:11px;border-radius:50%;'
            f'margin:0 2px;background:{"#55A868" if _b[i] else "#C44E52"}"></span>' for i in range(_nb))
        _s += "&nbsp;"
        _s += "".join(
            f'<span title="edge {i + 1}" style="display:inline-block;width:11px;height:11px;border-radius:50%;'
            f'margin:0 2px;border:2px solid {"#55A868" if _b[_nb + i] else "#C44E52"};box-sizing:border-box"></span>'
            for i in range(_ne))
        return _s

    def _pre(text, border):
        return (f'<pre style="white-space:pre-wrap;font-size:12px;line-height:1.5;margin:6px 0;padding:8px 10px;'
                f'border-left:4px solid {border};background:rgba(127,127,127,.07);border-radius:0 8px 8px 0">'
                f"{html.escape(text)}</pre>")

    def _diff(a, b):
        if a == b:
            return ('<div style="font-size:12.5px;font-weight:700;color:#C44E52;margin:4px 0">'
                    "↺ 跟上一版一字不差——它把同一份程式又交了一次</div>")
        _lines = list(difflib.unified_diff(a.splitlines(), b.splitlines(), lineterm="", n=1))[2:]
        _out = []
        for _ln in _lines[:40]:
            _c = "#2E7D32" if _ln.startswith("+") else "#C62828" if _ln.startswith("-") else "#8a949b"
            _out.append(f'<span style="color:{_c}">{html.escape(_ln)}</span>')
        return ('<pre style="white-space:pre-wrap;font-size:12px;line-height:1.45;margin:6px 0;padding:8px 10px;'
                'background:rgba(127,127,127,.07);border-radius:8px">' + "\n".join(_out) + "</pre>")

    _cards = [
        '<div style="font-size:13px;line-height:1.7;margin-bottom:6px"><b>一句話需求：</b>'
        + html.escape(_p["prompt"]) + "</div>"
    ]
    for _i, _row in enumerate(_rows):
        _code = V["codes"][_row[0]]
        _ok = _row[1] + _row[2] == _nb + _ne
        _badge = ("✅ 完整測試全過" if _ok else
                  f"❌ 基本 {_row[1]}/{_nb}、邊界 {_row[2]}/{_ne}") + ("（逾時）" if _row[3] else "")
        _card = (f'<div style="border:2px solid {"#55A868" if _ok else "#9AA7AE"};border-radius:12px;'
                 f'padding:10px 12px;margin:10px 0">'
                 f'<div style="font-weight:800;font-size:13.5px">第 {_i + 1} 版　{_badge}</div>'
                 f'<div style="margin:6px 0">{_dots(_row)}</div>')
        if _i > 0:
            _card += '<div style="font-size:12px;color:#8a949b">跟上一版的 diff：</div>'
            _card += _diff(V["codes"][_rows[_i - 1][0]], _code)
        _card += '<details><summary style="cursor:pointer;font-size:12.5px">看完整程式</summary>' + _pre(_code, "#4C72B0") + "</details>"
        if _row[4] >= 0:
            _card += ('<div style="font-size:12px;color:#8a949b;margin-top:6px">AI 接著收到的回饋：</div>'
                      + _pre(V["outs"][_row[4]], "#DD8452"))
        elif trace_cond.value == "reroll" and not _ok and _i + 1 < len(_rows):
            _card += '<div style="font-size:12px;color:#8a949b;margin-top:6px">（重抽：沒有回饋，開新對話再生成一次）</div>'
        _card += "</div>"
        _cards.append(_card)
    _last = _rows[-1]
    if trace_cond.value == "weak" and _last[1] == _nb and _last[2] < _ne:
        _cards.append('<div style="border-left:4px solid #C44E52;padding:8px 12px;font-size:13px;line-height:1.7">'
                      "<b>基本測試全綠，迴圈停了</b>——但空心的邊界測試還是紅的。AI 看不到的測試，它不會去修；"
                      "在它眼裡這一版已經「做完了」。</div>")
    mo.Html("<div>" + "".join(_cards) + "</div>")
    return


@app.cell(hide_code=True)
def _(mo):
    mo.md(
        r"""
    ## 3️⃣ 你是測試設計師：測試寫到哪，AI 就看到哪

    前面 8 次重跑 × 5 種下一步，這一題累積了一批 AI 寫的程式（一字不差的只算一份）。
    現在**由你決定測試集**（預設只放基本測試）：從清單加入或移除測試案例，看有多少份程式「被你的測試放行、其實是錯的」。

    這一格是**當場在你的瀏覽器裡**把每一份程式重跑一遍——不是查表。
    """
    )
    return


@app.cell
def _(V, mo):
    _names = [p["name"] for p in V["problems"]]
    design_prob = mo.ui.dropdown(options=_names, value="format_bytes", label="題目")
    design_prob
    return (design_prob,)


@app.cell
def _(PROB, design_prob, mo):
    _p = PROB[design_prob.value]

    def _label(kind, i, args, exp):
        _call = f"{design_prob.value}({', '.join(repr(a) for a in args)})"
        _rhs = f"raise {exp['raises']}" if isinstance(exp, dict) else repr(exp)
        return f"{'基本' if kind == 'basic' else '邊界'} {i}：{_call} → {_rhs}"

    _opts = {}
    for _kind in ("basic", "edge"):
        for _i, (_args, _exp) in enumerate(_p[_kind], 1):
            _opts[_label(_kind, _i, _args, _exp)] = (_kind, _i - 1)
    _basic_only = [k for k, v in _opts.items() if v[0] == "basic"]
    design_tests = mo.ui.multiselect(options=_opts, value=_basic_only, label="你的測試集")
    design_tests
    return (design_tests,)


@app.cell
def _(PROB, V, check_one, design_prob, safe_exec):
    # 這一題所有 AI 寫過的程式（去重、排除當時跑到逾時的），逐一在瀏覽器裡重跑全部測試案例
    _p = PROB[design_prob.value]
    _cases = [("basic", i, a, e) for i, (a, e) in enumerate(_p["basic"])] + \
             [("edge", i, a, e) for i, (a, e) in enumerate(_p["edge"])]
    _ids = sorted({row[0] for t in V["tasks"] if t["p"] == design_prob.value
                   for rows in t["b"].values() for row in rows if not row[3]})
    design_results = []
    for _cid in _ids:
        _fn = safe_exec(V["codes"][_cid], design_prob.value)
        design_results.append((_cid, {(k, i): check_one(_fn, a, e) for k, i, a, e in _cases}))
    return (design_results,)


@app.cell
def _(design_results, design_tests, plt):
    _sel = list(design_tests.value)
    _acc_ok = _acc_bad = _rej = 0
    for _cid, _res in design_results:
        _all_ok = all(_res.values())
        _accepted = all(_res[s] for s in _sel)
        if _accepted and _all_ok:
            _acc_ok += 1
        elif _accepted:
            _acc_bad += 1
        else:
            _rej += 1
    _fig, _ax = plt.subplots(figsize=(6.4, 2.5))
    _left = 0
    _handles = []
    for _v, _c, _nm in [(_acc_ok, "#55A868", "passed & truly correct"),
                        (_acc_bad, "#C44E52", "passed YOUR tests but wrong"),
                        (_rej, "#9AA7AE", "caught by your tests")]:
        _h = _ax.barh([0], [_v], left=_left, color=_c, edgecolor="#1C2B33", linewidth=1, height=0.5,
                      label=f"{_nm}: {_v}")
        _handles.append(_h)
        if _v:
            _ax.text(_left + _v / 2, 0, str(_v), ha="center", va="center", color="white", fontweight="bold")
        _left += _v
    _ax.set_xlim(0, max(1, _left))
    _ax.set_yticks([])
    _ax.xaxis.set_major_locator(plt.MaxNLocator(integer=True))
    _ax.set_xlabel(f"unique AI-written programs for this problem ({_left})")
    _fig.legend(handles=_handles, fontsize=8.5, loc="lower center", ncol=1, frameon=False)
    _fig.subplots_adjust(left=0.04, right=0.97, top=0.88, bottom=0.56)
    _ax.set_title(f"your test suite: {len(_sel)} case(s)", fontsize=10)
    _fig
    return


@app.cell
def _(PROB, V, design_prob, design_results, design_tests, html, mo):
    _sel = list(design_tests.value)
    _p = PROB[design_prob.value]
    _bad = [(c, r) for c, r in design_results if all(r[s] for s in _sel) and not all(r.values())]
    if not _sel:
        _msg = "你的測試集是空的——**每一份程式都會被放行**。沒有測試，就沒有眼睛。"
        _ex = ""
    elif not _bad:
        _msg = "你的測試集沒有放行任何錯誤的程式。🎯 試試看拿掉幾個邊界案例，看破口怎麼出現。"
        _ex = ""
    else:
        _cid, _res = _bad[0]
        _missed = [k for k, v in _res.items() if not v]
        _k, _i = _missed[0]
        _args, _exp = _p[_k][_i]
        _call = f"{design_prob.value}({', '.join(repr(a) for a in _args)})"
        _want = f"raise {_exp['raises']}" if isinstance(_exp, dict) else repr(_exp)
        _msg = (f"有 **{len(_bad)} 份**錯誤的程式會被你的測試集放行。舉一份為例——它在你沒放進來的 "
                f"`{_call}` 上會出錯（應該是 `{_want}`）：")
        _ex = ('<pre style="white-space:pre-wrap;font-size:12px;line-height:1.5;padding:8px 10px;'
               'border-left:4px solid #C44E52;background:rgba(127,127,127,.07);border-radius:0 8px 8px 0">'
               + html.escape(V["codes"][_cid]) + "</pre>")
    mo.vstack([mo.md(_msg), mo.Html(_ex)])
    return


@app.cell(hide_code=True)
def _(RISK, mo):
    mo.md(
        rf"""
    ## 4️⃣ 兩顆地雷：不存在的套件、寫死的金鑰

    測試抓得到「答錯」，抓不到「裝了什麼」和「把什麼寫進程式碼」。兩個實測（{RISK["meta"]["model"]}，{RISK["meta"]["date"]}）：

    **A. 套件幻覺**：請模型為 15 個任務各推薦 3 個 pip 套件（每題問 {RISK["meta"]["samples"]} 次），
    再把每個名字**真的拿去查 PyPI**（`https://pypi.org/pypi/<名字>/json`：200＝存在、404＝不存在）。
    不存在的名字，就是攻擊者可以搶先註冊、等你 `pip install` 的空位（slopsquatting）。
    """
    )
    return


@app.cell
def _(RISK, mo):
    _tasks = list(dict.fromkeys(r["task"] for r in RISK["pkg"]))
    pkg_task = mo.ui.dropdown(options=_tasks, value=_tasks[6], label="任務")
    pkg_task
    return (pkg_task,)


@app.cell
def _(RISK, np, pkg_task, plt):
    _tasks = list(dict.fromkeys(r["task"] for r in RISK["pkg"]))
    _kind = {r["task"]: r["kind"] for r in RISK["pkg"]}
    _rate, _cnt = [], []
    for _tk in _tasks:
        _all = [c for r in RISK["pkg"] if r["task"] == _tk for c in r["packages"]]
        _miss = sum(not c["exists"] for c in _all)
        _rate.append(_miss / max(1, len(_all)))
        _cnt.append(f"{_miss}/{len(_all)}")
    _fig, _ax = plt.subplots(figsize=(6.4, 4.4))
    _y = np.arange(len(_tasks))[::-1]
    _cols = ["#4C72B0" if _kind[t] == "mainstream" else "#DD8452" for t in _tasks]
    _ax.barh(_y, _rate, color=_cols)
    for _yy, _r, _t in zip(_y, _rate, _cnt):
        _ax.text(_r + 0.01, _yy, _t, va="center", fontsize=8.5, color="#1C2B33")
    _ax.set_yticks(_y)
    _ax.set_yticklabels([f"task {i + 1}" + (" (common)" if _kind[t] == "mainstream" else " (local/niche)")
                         for i, t in enumerate(_tasks)], fontsize=8.5)
    for _lab, _t in zip(_ax.get_yticklabels(), _tasks):
        if _t == pkg_task.value:
            _lab.set_fontweight("bold")
            _lab.set_color("#C44E52")
    _ax.set_xlim(0, 1.1)
    _ax.set_xlabel("share of recommended names NOT on PyPI (404)")
    _ax.set_title("package hallucination by task (red label = your pick)", fontsize=10)
    _ax.grid(axis="x", alpha=0.3)
    _fig.tight_layout()
    _fig
    return


@app.cell
def _(RISK, html, mo, pkg_task):
    _tasks = list(dict.fromkeys(r["task"] for r in RISK["pkg"]))
    _rows = [r for r in RISK["pkg"] if r["task"] == pkg_task.value]
    _cells = []
    for _r in _rows:
        _items = []
        for _c in _r["packages"]:
            if _c["exists"]:
                _items.append(f'✅ <code>{html.escape(_c["name"])}</code> <span style="color:#8a949b">'
                              f'（PyPI 有：{html.escape(_c["version"])}，最後發版 {_c["last_upload"]}）</span>')
            else:
                _items.append(f'❌ <code>{html.escape(_c["name"])}</code> <span style="color:#C44E52">'
                              "（PyPI 查無此套件：404）</span>")
        _cells.append(f'<div style="margin:6px 0"><b>第 {_r["sample"] + 1} 次問</b><br>' + "<br>".join(_items) + "</div>")
    _i = _tasks.index(pkg_task.value) + 1
    mo.Html(
        f'<div style="font-size:13px;line-height:1.8"><div>task {_i}：「{html.escape(pkg_task.value)}」</div>'
        + "".join(_cells) + "</div>"
    )
    return


@app.cell
def _(RISK, html, mo):
    _sec = RISK["secrets"]
    _tally = {k: sum(r["where"] == k for r in _sec) for k in ("hardcoded", "env", "none")}
    _old_api = sum("ChatCompletion" in r["code"] for r in _sec)
    _ex = next((r for r in _sec if r["where"] == "hardcoded"), None)
    _code = ""
    if _ex:
        _code = ('<pre style="white-space:pre-wrap;font-size:12px;line-height:1.5;padding:8px 10px;'
                 'border-left:4px solid #C44E52;background:rgba(127,127,127,.07);border-radius:0 8px 8px 0">'
                 + html.escape("\n".join(ln for ln in _ex["code"].splitlines() if not ln.startswith("```"))[:900])
                 + "</pre>")
    mo.vstack([
        mo.md(
            f"""
    **B. 寫死金鑰**：請它寫「呼叫 OpenAI API 把文字摘要成一句話」的函式 {len(_sec)} 次，看 API key 放哪：

    | 金鑰放哪 | 次數 |
    |---|---|
    | 寫死在程式碼裡的字串（例如 `api_key="..."`） | **{_tally["hardcoded"]}** |
    | 從環境變數讀（`os.environ`／`os.getenv`） | {_tally["env"]} |
    | 沒處理（交給 SDK 預設行為） | {_tally["none"]} |

    寫死的字串一旦 commit 就進了 git 歷史——就算下一版刪掉，舊版還在。
    另外，這 {len(_sec)} 份裡有 **{_old_api}** 份用的是 `openai.ChatCompletion.create`——openai 套件 1.0 版就移除的舊寫法，
    在今天的 SDK 上一呼叫就丟 `APIRemovedInV1`（2026-09 用 openai 3.19.2 實測）——這種錯不用連網就會炸，任何一個呼叫到它的測試都抓得到。
    """
        ),
        mo.Html(_code),
    ])
    return


@app.cell(hide_code=True)
def _(mo):
    mo.md(
        r"""
    ## 5️⃣ 你的實驗區：回饋預算怎麼花最划算

    每多試一次都要付時間與 token。下面用**實測的**每次嘗試耗時與輸出 token，
    幫你算「一種策略、試到第幾次」的成績與代價。挑戰題在教學頁的「換你動手」。
    """
    )
    return


@app.cell
def _(COND_LABEL, CONDS, mo):
    my_cond = mo.ui.dropdown(
        options={COND_LABEL[c]: c for c in CONDS},
        value=COND_LABEL["full"], label="策略",
    )
    my_k = mo.ui.slider(1, 4, 1, value=4, label="最多試幾次", show_value=True)
    mo.hstack([my_cond, my_k], justify="start", gap=2, wrap=True)
    return my_cond, my_k


@app.cell
def _(COND_COLOR, COND_EN, CONDS, MAXK, RUNS, V, att_at, is_pass, my_cond, my_k, np, plt):
    def _stats(c, k):
        _solved = np.zeros(RUNS)
        _tok = np.zeros(RUNS)
        _sec = np.zeros(RUNS)
        for _t in V["tasks"]:
            _rows = _t["b"][c][:k]
            _solved[_t["r"]] += is_pass(_t["p"], att_at(_rows, k))
            _tok[_t["r"]] += sum(r[6] for r in _rows)
            _sec[_t["r"]] += sum(r[5] for r in _rows)
        return _solved, _tok, _sec

    budget = {(c, k): _stats(c, k) for c in CONDS for k in range(1, MAXK + 1)}
    _fig, _ax = plt.subplots(figsize=(6.4, 3.8))
    for _c in CONDS:
        _xs = [budget[(_c, k)][1].mean() for k in range(1, MAXK + 1)]
        _ys = [budget[(_c, k)][0].mean() for k in range(1, MAXK + 1)]
        _ax.plot(_xs, _ys, "-o", color=COND_COLOR[_c], label=COND_EN[_c], linewidth=2, markersize=4)
    _s, _t2, _ = budget[(my_cond.value, my_k.value)]
    _ax.plot([_t2.mean()], [_s.mean()], "o", markersize=15, markerfacecolor="none",
             markeredgecolor="#1C2B33", markeredgewidth=2.2)
    _ax.set_xlabel("output tokens spent per run of 12 problems (mean)")
    _ax.set_ylabel("problems passing all tests (mean)")
    _ax.set_title("score vs. token budget (circle = your choice)", fontsize=10)
    _ax.grid(alpha=0.3)
    _ax.legend(fontsize=8.5, loc="lower right")
    _fig.tight_layout()
    _fig
    return (budget,)


@app.cell
def _(COND_LABEL, NPROB, budget, mo, my_cond, my_k):
    _s, _tok, _sec = budget[(my_cond.value, my_k.value)]
    _s1, _tok1, _sec1 = budget[(my_cond.value, 1)]
    _gain = _s.mean() - _s1.mean()
    _extra = _tok.mean() - _tok1.mean()
    _per = f"每多救回 1 題約花 **{_extra / _gain:,.0f}** 個輸出 token" if _gain > 0.05 else "多試的次數幾乎沒有救回任何題目"
    mo.md(
        f"""
    **{COND_LABEL[my_cond.value]}、最多試 {my_k.value} 次**：平均 **{_s.mean():.1f} / {NPROB}** 題全過
    （最少 {_s.min():.0f}、最多 {_s.max():.0f}），每輪 12 題平均花 **{_tok.mean():,.0f}** 個輸出 token、
    生成時間合計約 **{_sec.mean():.0f} 秒**。

    跟只試 1 次比：多救回 {_gain:.1f} 題，{_per}。
    """
    )
    return


@app.cell(hide_code=True)
def _(mo):
    mo.accordion(
        {
            "💡 LEVEL 1 參考解答": mo.md(
                r"""
    1️⃣ 的表格最後一欄（2026-09 這批紀錄）：只說「不對」72%、貼錯誤原文 67%、錯誤原文＋一句診斷 42%、
    開新對話重抽 14%。**給的資訊越具體，它越不會原樣照交**；而重抽是全新對話，沒有「上一版」可以抄。
    在 2️⃣ 選 `round_half_up`、第 5 次重跑：「只說不對」三版都是 `int(x + 0.5)`（↺ 一字不差）；
    「貼回錯誤原文」這次第 2 版就改對了——但 8 次重跑裡只有 3 次這麼幸運。
    回饋裡明明寫著 `assert -2 == -3`，它還是可能原封不動：**看得見不等於看得懂**。
    """
            ),
            "💡 LEVEL 2 參考解答": mo.md(
                r"""
    `format_bytes`：只留兩個基本測試時，被放行的程式裡有 **15 份是錯的、只有 1 份全對**。
    一次只加一個邊界案例：加 `1024**5 → '1024.0 TB'` 最有效（錯的剩 1 份）；
    加 `-1 → raise ValueError` 剩 4 份、`512 → '512 B'` 剩 5 份、`1024 → '1.0 KB'` 剩 12 份；四個全加，錯的歸零。
    `split_bill`：只留基本測試放行 7 份錯的；三個金額案例（例如 `(100, 3) → [34, 33, 33]`）**任何一個**都能全部擋下，
    只加 `n=0 要 raise` 那個則剩 5 份。
    **最有效的案例，是那個「你心裡有答案、但需求裡沒寫」的細節。**
    """
            ),
            "💡 LEVEL 3 提示": mo.md(
                r"""
    這批紀錄的走勢：「只說不對」與「貼錯誤原文」在第 2 次之後完全不動（2.75 題）；
    「一句診斷」第 2 次就從 2.0 跳到 4.4 題、第 3 次再多一點；「重抽」每次都還在漲，但一次只多一點。
    換算成「每多救回 1 題花幾個輸出 token」，一句診斷反而最省——資訊對了，模型少走冤枉路。
    （圖上只算輸出 token；同一段對話的輸入會越滾越長，重抽不會，所以真實帳單差距只會更大。）
    設計你的流程時，自我驗證的標準是：它要能同時解釋兩件事——
    (1) 為什麼「只有基本測試」那條線很便宜（很早就停手），(2) 為什麼便宜不代表好（回到 3️⃣ 看它放行了什麼）。
    """
            ),
        }
    )
    return


if __name__ == "__main__":
    app.run()
