前九章一直在讲「怎么让《合成小岛》跑起来、怎么变现」，第 10 章换一个视角：在把它推到真实平台之前，法律与平台规则先于玩法挡住你。本节只处理最上游的一件事——把「资质、备案、平台类目要求」整理成一张能被程序读进来做校验的清单，并给出可跑的覆盖率与 schema 对照脚本。具体该办哪些证、走什么流程、审核能不能过，需真实业务环境与专业法务判断，命令行不可复现（X），本节不提供任何法律意见或合规结论。
先说分工。站内已有一篇小游戏开发综述 ，那篇 96 行短文里有审核合规口径的一段综述，讲「要准备哪些材料」；本节不重复它的框架，而是把同一类材料变成一份带字段、带状态、可校验的配置——综述告诉你「要办证」，本节告诉你「证办到哪一步、缺哪一项、字段填得对不对」。上线流程本身（提审、灰度、回滚）是第 9 章的事，本节不越界，只把「清单何时补齐」作为交接点。
资质、备案、平台规则是三件事
先分清三种东西，它们的责任主体和补齐时点都不同。资质是主体层面的东西：营业执照、软件著作权登记，开发期就该办。备案是向监管与平台申报的东西：ICP 备案、含内购时的游戏版号、平台类目资质，提审前必须齐。平台规则是平台侧的准入与运营约束：类目要求、支付结算资质、未成年人保护接入，横跨提审与上线两个阶段。三者混在一起，最容易出的错是把「上线后才需要的」当成「提审前必须有」，白白拖慢排期；或者反过来，提审前漏了「含内购才需要」的版号。
本节的工程化思路很朴素：把每一项做成一条带 stage（最迟补齐阶段）、level（must / must-if-paid / should）和 category 的记录，再用一段脚本对「当前已持有集合」算覆盖率。must-if-paid 是关键——它的必要性取决于游戏有没有内购，所以配置里带一个 hasPaid 开关，有内购时才把它计入必办清单。
把清单写成可校验的数据（D 层）
下面这段是 D 层静态配置：资质条目、三阶段的「已持有集合」、备案项的 schema 与样例。所有主体名、编号、状态都是示例占位；真实资质要求与审核结果需真实业务环境与专业法务判断，命令行不可复现（X）。
{
"version": 1,
"game": "merge-island",
"observedAt": "2026-10",
"note": "资质、备案与平台规则的检查清单（D 层）。条目是对公开可观察机制的归纳，用于工程侧做「清单覆盖率 / 备案项 schema」两类可复现校验；真实主体资质要求、真实备案流程与真实审核结果需真实业务环境与专业法务判断，命令行不可复现（X）。所有编号、主体名与状态均为示例占位。",
"stages": ["dev", "prereview", "live"],
"hasPaid": true,
"items": [
{"id": "entity-license", "name": "主体营业执照", "stage": "dev", "level": "must", "category": "subject"},
{"id": "soft-copyright", "name": "软件著作权登记", "stage": "dev", "level": "must", "category": "ip"},
{"id": "icp-filing", "name": "ICP 备案（域名/主体）", "stage": "prereview", "level": "must", "category": "filing"},
{"id": "game-approval", "name": "游戏版号（含内购时）", "stage": "prereview", "level": "must-if-paid", "category": "filing"},
{"id": "category-permit", "name": "平台类目资质", "stage": "prereview", "level": "must", "category": "platform"},
{"id": "privacy-policy", "name": "隐私政策与用户协议", "stage": "prereview", "level": "must", "category": "privacy"},
{"id": "realname-antiaddiction", "name": "实名与防沉迷接入", "stage": "prereview", "level": "must", "category": "minor"},
{"id": "pay-qualify", "name": "支付 / 结算资质", "stage": "live", "level": "must-if-paid", "category": "payment"},
{"id": "minor-protection-report", "name": "未成年人保护专项报告", "stage": "live", "level": "should", "category": "minor"}
],
"holdings": {
"dev": ["entity-license"],
"prereview": ["entity-license", "soft-copyright", "icp-filing"],
"live": ["entity-license", "soft-copyright", "icp-filing", "game-approval", "category-permit", "privacy-policy", "realname-antiaddiction", "pay-qualify"]
},
"filingSchema": {
"required": ["id", "title", "entity", "category", "status", "evidence"],
"statusEnum": ["pending", "submitted", "approved", "rejected"],
"evidenceRequiredUnless": ["pending"]
},
"filingItems": [
{"id": "icp-filing", "title": "ICP 备案", "entity": "示例科技（深圳）有限公司", "category": "filing", "status": "submitted", "evidence": "示例-ICP-0001"},
{"id": "game-approval", "title": "游戏版号", "entity": "示例科技（深圳）有限公司", "category": "filing", "status": "pending", "evidence": ""},
{"id": "pay-qualify", "title": "支付资质", "entity": "", "category": "payment", "status": "approved", "evidence": ""}
]
}
用 S 层脚本做覆盖率与 schema 对照
下面这段是 S 层仿真脚本，读上面那份配置，做两件事：一是按「已持有集合」算每个阶段的必办清单覆盖率（required = must 加上有内购时的 must-if-paid，should 不进分母），列出缺失项与最迟补齐阶段；二是对备案项做 schema 校验——必填字段是否齐、status 是否落在枚举内、非 pending 状态的凭证号是否为空。产出 JSON + CSV + SVG，三阶段一对照，「开发期缺得多、上线后应齐」应当一眼可见。
#!/usr/bin/env python3
# file: sim/l10_license.py
# tier: S
# verify: /opt/homebrew/bin/python3 sim/l10_license.py --config config/ --out out/
"""资质备案检查：清单覆盖率 + 备案项 schema 校验（整数、可复算）。
读 config/l10_license.json：资质条目 + 各阶段「已持有」集合 + 备案项 schema 与样例。
1) 覆盖率：required = level 以 "must" 开头（含 must-if-paid，当 hasPaid 为真时计入）；
coveragePermille = |held ∩ required| * 1000 // |required|；
2) 缺失清单：required - held，按条目 stage 标注「最迟补齐阶段」；
3) 备案项 schema 校验：required 字段齐全、status 落在枚举内、
status 不在 evidenceRequiredUnless 时 evidence 非空；
4) 三阶段对照（dev / prereview / live）——同一份清单，只换「已持有」集合。
全程整数，比例用千分比。产出 out/l10_license.json + out/l10_license.csv + out/l10_license.svg。
清单条目与 schema 均为示例；真实资质要求、真实备案流程与审核结果需真实业务环境与
专业法务判断，命令行不可复现（X）。
"""
from __future__ import annotations
import argparse
import csv
import json
import os
import sys
import xml.etree.ElementTree as ET
STAGE_LABEL = {"dev": "开发期", "prereview": "提审前", "live": "上线后"}
def load(cfg_dir: str, name: str) -> dict:
with open(os.path.join(cfg_dir, f"{name}.json"), encoding="utf-8") as f:
return json.load(f)
def required_ids(cfg: dict) -> list[str]:
"""required：must / must-if-paid（有内购时计入）。should 不计入覆盖率分母。"""
out = []
for it in cfg["items"]:
if it["level"] == "must" or (it["level"] == "must-if-paid" and cfg["hasPaid"]):
out.append(it["id"])
return out
def coverage(cfg: dict, stage: str) -> dict:
req = required_ids(cfg)
held = set(cfg["holdings"][stage])
got = [i for i in req if i in held]
missing = [i for i in req if i not in held]
return {"stage": stage, "required": len(req), "held": len(got),
"coveragePermille": len(got) * 1000 // len(req),
"missing": missing}
def validate_filing(cfg: dict) -> list[dict]:
"""备案项 schema 校验：返回违规列表（itemId + 字段 + 原因）。"""
schema = cfg["filingSchema"]
enum = set(schema["statusEnum"])
allow_empty = set(schema["evidenceRequiredUnless"])
bad = []
for row in cfg["filingItems"]:
iid = row.get("id", "?")
for key in schema["required"]:
if key not in row:
bad.append({"itemId": iid, "field": key, "reason": "missingField"})
if row.get("status") not in enum:
bad.append({"itemId": iid, "field": "status", "reason": "badEnum"})
if row.get("status") not in allow_empty and not str(row.get("evidence", "")).strip():
bad.append({"itemId": iid, "field": "evidence", "reason": "emptyEvidence"})
if not str(row.get("entity", "")).strip():
bad.append({"itemId": iid, "field": "entity", "reason": "emptyEntity"})
return bad
def txt(parent: ET.Element, x: float, y: float, s: str, size: int = 12,
fill: str = "#374151", anchor: str | None = None) -> None:
a = {"x": str(x), "y": str(y), "font-size": str(size), "fill": fill}
if anchor:
a["text-anchor"] = anchor
ET.SubElement(parent, "text", a).text = s
def svg(path: str, rows: list[dict]) -> int:
"""纯 Python SVG：三阶段覆盖率横条。"""
W, PAD, ROW = 720, 120, 46
H = 70 + ROW * len(rows)
r = ET.Element("svg", {"xmlns": "http://www.w3.org/2000/svg", "width": str(W),
"height": str(H), "viewBox": f"0 0 {W} {H}"})
ET.SubElement(r, "rect", {"width": str(W), "height": str(H), "fill": "#ffffff"})
txt(r, 16, 26, "《合成小岛》资质清单覆盖率（示例清单，‰）", 15, "#111827")
bmax = W - PAD - 90
for i, row in enumerate(rows):
y = 70 + i * ROW
txt(r, 16, y + 20, f'{STAGE_LABEL[row["stage"]]} ({row["held"]}/{row["required"]})')
w = bmax * row["coveragePermille"] // 1000
color = "#dc2626" if row["coveragePermille"] < 500 else (
"#f59e0b" if row["coveragePermille"] < 1000 else "#10b981")
ET.SubElement(r, "rect", {"x": str(PAD), "y": str(y + 6), "width": str(w),
"height": "20", "rx": "4", "fill": color})
txt(r, PAD + w + 8, y + 21, f'{row["coveragePermille"]}‰', 12, "#6b7280")
os.makedirs(os.path.dirname(path), exist_ok=True)
ET.ElementTree(r).write(path, encoding="utf-8", xml_declaration=True)
return os.path.getsize(path)
def main() -> int:
ap = argparse.ArgumentParser()
ap.add_argument("--config", default="config/")
ap.add_argument("--out", default="out/")
a = ap.parse_args()
cfg = load(a.config, "l10_license")
name_of = {it["id"]: it["name"] for it in cfg["items"]}
stage_of = {it["id"]: it["stage"] for it in cfg["items"]}
rows = [coverage(cfg, s) for s in cfg["stages"]]
for row in rows:
row["missingDetail"] = [
{"id": m, "name": name_of[m], "dueStage": stage_of[m]} for m in row["missing"]]
violations = validate_filing(cfg)
os.makedirs(a.out, exist_ok=True)
sb = svg(os.path.join(a.out, "l10_license.svg"), rows)
report = {"source": "sim/l10_license.py", "tier": "S", "game": cfg["game"],
"observedAt": cfg["observedAt"], "note": cfg["note"],
"requiredCount": len(required_ids(cfg)), "hasPaid": cfg["hasPaid"],
"stages": rows, "filingViolations": violations, "svgBytes": sb}
with open(os.path.join(a.out, "l10_license.json"), "w", encoding="utf-8") as f:
json.dump(report, f, ensure_ascii=False, indent=2)
with open(os.path.join(a.out, "l10_license.csv"), "w", encoding="utf-8", newline="") as f:
w = csv.writer(f)
w.writerow(["stage", "required", "held", "coveragePermille", "missing"])
for row in rows:
w.writerow([row["stage"], row["required"], row["held"],
row["coveragePermille"], "|".join(row["missing"])])
for row in rows:
print(f'{STAGE_LABEL[row["stage"]]:<4} 覆盖 {row["held"]}/{row["required"]} '
f'= {row["coveragePermille"]}‰ 缺: {",".join(row["missing"]) or "无"}')
print(f"备案项 schema 违规 {len(violations)} 条:")
for v in violations:
print(f' {v["itemId"]}.{v["field"]} -> {v["reason"]}')
return 0
if __name__ == "__main__":
sys.exit(main())
本机跑出来的输出（同机相对量级，绝对数值不可移植）：
开发期 覆盖 1/8 = 125‰ 缺: soft-copyright,icp-filing,game-approval,category-permit,privacy-policy,realname-antiaddiction,pay-qualify
提审前 覆盖 3/8 = 375‰ 缺: game-approval,category-permit,privacy-policy,realname-antiaddiction,pay-qualify
上线后 覆盖 8/8 = 1000‰ 缺: 无
备案项 schema 违规 2 条:
pay-qualify.evidence -> emptyEvidence
pay-qualify.entity -> emptyEntity
三阶段的覆盖率对照（125‰ → 375‰ → 1000‰）把「清单何时补齐」变成一条可见的曲线。更值得注意的是最后两条 schema 违规：pay-qualify 这条记录主体名空、且 approved 状态却没有凭证号——它在覆盖率里被算作「已持有」，但字段是残缺的。这正是覆盖率会骗人的地方：「有没有办」和「办得对不对」是两件事，只有 schema 校验能把后者揪出来。
平台规则的引用纪律（X）
上面所有条目都是对公开可观察机制的归纳，观察时点为 2026-10。真实平台上，具体要办哪些证、类目资质如何核验、版号审批要多久，都属于不可复现的真实业务过程，命令行跑不出来，标 X。本节的做法不是给你一份「标准答案清单」，而是给你一套装清单的容器：把真实法务给出的结论填进 items 与 filingSchema，脚本负责校验一致性与完整性。任何具体资质结论都必须由专业法务判断，脚本只保证「清单内部自洽、字段不残缺」。
把清单接到上线流程（交接点）
本节只负责把清单做对，不负责流程编排。清单里每一条的 stage 字段就是与第 9 章上线流程的交接点：开发期该办的在开工前办、提审前该办的在提审材料冻结前办、上线后该办的在上线后一个月内补齐。第 9 章讲提审、灰度与回滚的流程本身，本节讲清单的内容与校验，两者在「提审材料冻结」这一刻对接，不重叠。
阅读导航：上一节：目录 · 下一节：10.2 内容安全与审核红线 。
继续阅读
探索更多技术文章
浏览归档，发现更多关于系统设计、工具链和工程实践的内容。