v0.5.115: add History routing and SEO prerender for Climperor Web.

Path URLs, crawlable hero/mechanics pages, and sitemap make the static site indexable while keeping SPA hydration. Co-authored-by: Cursor <cursoragent@cursor.com>
2026-07-30 04:55:33 +08:00
parent 38f46ad2ea
commit 544ea42d40
16 changed files with 838 additions and 81 deletions
@@ -70,8 +70,9 @@ climperor/
 | `web/fetch_patches.py` | 拉取近一年（默认 365 天，`--days`/`--since`）版本列表 + 逐版本 `patchnotes` 详情 → `web/data/patches.json`；构建 id→名称/图标的 `lookup` 并下载引用到的物品/技能图标（`--no-icons` 跳过；`--force` 重抓全量；`--check` 只比对列表与本地 details，stdout JSON） |
 | `web/requirements.txt` | Web 刷新依赖（`oss2`）；Gitea Actions 仅装它 |
 | `web/fetch_stratz_meta.py` | 拉取 STRATZ 各勋章段位 `winWeek`（近 N 周 pick/win + 同段位最近 1 周分路，`positionIds`）+ 对位 Top → `web/data/stratz_hero_meta.json` / `web/data/stratz_matchup_tops.json`（需 token；**仅上分帝 Web**；勿进 recommend / relations）。对位为**全局聚合**（无段位/分路/周过滤）；weekly 默认全量刷新，`--resume-matchups` 仅中断续跑；失败保留旧值并标 `stale` |
-| `web/serve_relations.py` | 上分帝 Web 本地开发服务（`web/frontend/`；改 `web/data/*.json` 后刷新；`/api/live-status` 调用 `fetch_streamer_live.probe_streamers` 做真实探测，内存缓存 60s，失败标 `stale`/不显示直播角标；`/streamer-video/` 提供主播高光 mp4 与同名 JPG 封面，支持 HTTP Range） |
-| `web/export_relations_site.py` | 导出上分帝 Web 为纯静态站点 → `web/dist/relations/`（data.json 快照 + 前端 + 图片；`SITE_VERSION` 常量与 `web/frontend/config.js` 同步；`--ability-video-base` / `--static-asset-base` 写 `config.js` 指向 OSS；设 `--static-asset-base` 时不拷贝图标进 dist；`--with-videos` 可选本地拷贝技能/主播视频，生产部署勿用） |
+| `web/serve_relations.py` | 上分帝 Web 本地开发服务（`web/frontend/`；改 `web/data/*.json` 后刷新；History 深度路径 SPA fallback 回 `index.html`；`/api/live-status` 调用 `fetch_streamer_live.probe_streamers` 做真实探测，内存缓存 60s，失败标 `stale`/不显示直播角标；`/streamer-video/` 提供主播高光 mp4 与同名 JPG 封面，支持 HTTP Range） |
+| `web/export_relations_site.py` | 导出上分帝 Web 为纯静态站点 → `web/dist/relations/`（data.json 快照 + 前端 + 图片；`SITE_VERSION` / `SITE_ORIGIN` 与 `web/frontend/config.js` 同步；`--ability-video-base` / `--static-asset-base` / `--site-origin` 写 `config.js`；设 `--static-asset-base` 时不拷贝图标进 dist；调用 `seo_prerender.py` 写英雄/机制预渲染 HTML + `sitemap.xml` / `llms.txt`；拷贝 `_redirects` / `robots.txt`；`--with-videos` 可选本地拷贝技能/主播视频，生产部署勿用） |
+| `web/seo_prerender.py` | 导出期 SEO/GEO：注入 title/description/canonical/OG/JSON-LD 与 `#seo-prerender` 正文；生成全英雄 `/heroes/{key}`、机制 `/mechanics/{effect}`、顶层页、`sitemap.xml`、`llms.txt` |
 | `web/deploy_relations.py` | 一键部署上分帝 Web 静态站点到 Cloudflare Pages（导出 + 资产预检 + `wrangler` 直传 + 绑域名；默认 OSS base 指向 `climperor` 桶的视频与静态图；凭据经 keyzoo 注入或 env） |
 | `web/refresh_web.py` | 上分帝 Web 数据分层刷新编排（`daily`/`weekly`/`patch`/`all`）；忽略纯时间戳与 `is_live` 的业务摘要有变更才 OSS upload + `deploy_relations.py`（亦监视 frontend / `relations.json` / `heroes.json` / 网格顺序）；写 `web/.refresh/summary.json`（步骤/耗时/stale/部署；`--dry-run` 只打印不写盘、不 restore/save 缓存）；同 runner 文件锁串行发布，patch 遇全量刷新则延后并标 `skipped`；`--skip-deploy` / `--skip-oss` / `--force-deploy`。`daily`：stats/排行/比赛/pro/主播/主播开播探测 + patch check；`weekly`：STRATZ meta + 物品 + counter；`patch`：仅 check，`has_new` 时详情 + abilities/商店/meta/fears |
 | `web/refresh_cache.py` | Gitea Actions 刷新状态缓存桥：文件锁内从 runner `~/.cache/climperor-web-refresh/` 恢复/保存，`web/.refresh-cache/` 供 Actions cache 冷启动备份；覆盖所有定时生成数据与 `patches.json`，主播只合并抓取字段，保留 git 中手工名单 |
@@ -110,7 +111,7 @@ climperor/
 | `web/data/pro_matches.json` | 明星选手近期联赛/锦标赛（可选天梯）对局（`by_pro`/`by_hero` + 终局出装/加点 + `lobby_type`/`origin`；由 `fetch_pro_matches.py` 生成；Web 顶层「比赛」侧栏筛「全部 / 职业 / 国服」+ 选手，与英雄详情「近期比赛」合并展示；进 daily；**不**参与局内推荐） |
 | `web/data/leaderboards.json` | Valve Immortal 四区 Top100（`china`/`europe`/`americas`/`se_asia`；仅排名+昵称等；由 `fetch_leaderboards.py` 生成；Web「排行」选手榜只读；**不**参与局内推荐） |
 | `web/data/streamers.json` | 主播目录（手工 `live_url`/profile URL + 常用英雄 + 可选精选视频 `video`/`video_title` + 可选 `video_poster`/`video_fit`/`video_crop`/`video_aspect`；抖音 profile 补全；平台含抖音 / B 站 / 斗鱼；`platform_meta`；Web 卡片：头像行 = 头像 \| 昵称+账号/获赞粉丝 \|「关注」，签名（`signature`）独立全宽行（最长 3 行）；在播时粉环 +「直播」角标叠在环底（抖音式 `bottom:-6px`，无间距）；列表排序：先 `is_live` 再粉丝数降序；视口分档滚播（远处不拉、近处 metadata、中部 `canplay` 且单路 `preload=auto`）+ 同名 JPG 封面占位；有 `live_url` 时点头像进直播间；**「直播」角标只信本轮成功探测**（线上 `/api/live-status` 边缘缓存 5 分钟；本地 `serve_relations` 同逻辑缓存 60s；`stale`/失败不显示角标；`data.json` daily 仅作首屏兜底直至接口返回；`live_url` 仅作点击入口）；由 `fetch_streamers.py` 补全；Web「主播」只读；**不**参与局内推荐） |
-| `web/data/stratz_hero_meta.json` | STRATZ 各勋章段位近 N 周 pick/win（`weeks`/`latest`）+ **同段位最近 1 周分路**（`winWeek`+`positionIds`）+ `meta_board`（由 `fetch_stratz_meta.py` 生成；Web 英雄详情「走势」优先 + 顶层「走势」`#/trends` 近 N 周榜；**不**参与局内推荐） |
+| `web/data/stratz_hero_meta.json` | STRATZ 各勋章段位近 N 周 pick/win（`weeks`/`latest`）+ **同段位最近 1 周分路**（`winWeek`+`positionIds`）+ `meta_board`（由 `fetch_stratz_meta.py` 生成；Web 英雄详情「走势」优先 + 顶层「走势」`/trends` 近 N 周榜；**不**参与局内推荐） |
 | `web/data/stratz_matchup_tops.json` | STRATZ 对位/协同 Top（counters/countered/synergies；由 `fetch_stratz_meta.py` 生成；全局聚合 + `scope`/`stale`/`fetched_at`；Web「对位」Tab；与定性 `relations.json` 分开展示；**不**参与局内推荐） |
 | `web/data/item_shop.json` | 商店 11 列目录（官网 basic/upgrade；由 `fetch_item_shop.py` 生成；物品页只读） |
 | `web/data/items_meta.json` | 成品装备描述与机制标签（由 `fetch_items_meta.py` 生成） |
@@ -120,8 +121,8 @@ climperor/
 | `web/data/hero_abilities.json` | 英雄技能与机制汇总（由 `fetch_hero_abilities.py` 生成；含 ability `tags`） |
 | `web/data/item_counter_stats.json` | OpenDota 对阵装备观测证据缓存（对阵购买率/条件胜率减同装备全局基线；可再生成；不进 recommend / relations） |
 | `web/data/hero_item_fears.json` | 英雄怕的装备（规则推导；Web「怕」行） |
-| `web/frontend/` | 上分帝 Web 前端静态资源（`index.html` / `config.js` / `app.js` / `style.css` / `router.js` / `mobile-gate.js` / `functions/`）；`config.js` 含 `SITE_VERSION`、`ABILITY_VIDEO_BASE`、`STATIC_ASSET_BASE`；版本页底部显示 `v{SITE_VERSION}`；技能演示按官网 16:9（有空间加宽至约 720px，`contain` 不裁左右）；移动端由 `mobile-gate.js` 拦截 |
-| `web/frontend/router.js` | Hash 路由：`parseHash` / `serializeHash` / `installRouter` / `syncStateToUrl`；状态↔URL 双向同步（顶层标签 `heroes\|rankings\|streamers\|matches\|trends\|mechanics\|items\|patches` / 英雄 + 子标签 `skills\|core\|fears\|trends\|matchups\|matches\|streamers\|patches` / Immortal 地区 `#/rankings[/region]` / 明星比赛 `#/matches[/account_id][?origin=pro\|china][&page=N]`（默认全部类型、每页 20 场） / 主播目录 `#/streamers` / 近 8 周走势榜 `#/trends[/bracket][?sort=pr]`（默认按胜率） / 机制查询 `#/mechanics[/{effect}]`（默认 `basic_dispel`） / 物品 / 版本 / 标签筛选 / 搜索；旧 `stats` / `#/rankings/meta` 别名兼容） |
+| `web/frontend/` | 上分帝 Web 前端静态资源（`index.html` / `config.js` / `app.js` / `style.css` / `router.js` / `mobile-gate.js` / `_redirects` / `robots.txt` / `functions/`）；`config.js` 含 `SITE_VERSION`、`SITE_ORIGIN`、`ABILITY_VIDEO_BASE`、`STATIC_ASSET_BASE`；版本页底部显示 `v{SITE_VERSION}`；技能演示按官网 16:9（有空间加宽至约 720px，`contain` 不裁左右）；移动端由 `mobile-gate.js` 拦截（搜索/AI 爬虫 UA 跳过） |
+| `web/frontend/router.js` | History 路径路由：`parseHash` / `serializeHash`（操作 pathname+search）/ `installRouter` / `syncStateToUrl`；状态↔URL 双向同步（顶层 `/heroes\|rankings\|streamers\|matches\|trends\|mechanics\|items\|patches` / 英雄 + 子标签 `skills\|core\|fears\|trends\|matchups\|matches\|streamers\|patches` / Immortal `/rankings[/region]` / 明星比赛 `/matches[/account_id][?origin=pro\|china][&page=N]` / 主播 `/streamers` / 走势 `/trends[/bracket][?sort=pr]` / 机制 `/mechanics[/{effect}]`（默认 `basic_dispel`） / 物品 / 版本 / 标签筛选 / 搜索；旧 `stats` / `/rankings/meta` 与 hash `#/...` 兼容） |
 | `web/assets/hero_portraits/` | 官网横版头像（上分帝 Web；默认 wide 面部构图，非匹配模板） |
 | `web/assets/attr_icons/` | 官网主属性图标（力量/敏捷/智力/全才，上分帝 Web 用） |
 | `web/assets/role_icons/` | Valve 选人定位筛选图标（透明 PNG；英雄页定位栏；本地 `/role-icon/`，线上 OSS `role-icon/`） |
@@ -163,9 +164,9 @@ Dota 2 GSI → pc/gsi_watch.py (:3223)

 - **合规**：仅 GSI + 屏幕截图；**禁止**读进程内存、注入、绕过 VAC。
 - **GSI 范围**：普通玩家视角拿不到双方 pick；GSI 只作阶段触发、`team_slot`、本人 `hero`、本机 `accountid`/`steamid`。本人顶栏槽位**只信** GSI，不用截屏名字亮度猜测。
- **克制 / 搭档数据**：机制克制/搭档只用定性边存 `shared/data/relations.json`（克制有向 + 搭档无向 + 理由），**不要**用胜率/场次表达机制克制，也**不要**写入 `shared/data/heroes.json`。Web 英雄页三视图：克制 / 被克制 / 搭档。OpenDota 段位胜率/场次单独存 `web/data/hero_stats.json`（近约 7 天窗；上场率 = 出场/(Σ出场/10)；冠绝样本过小时与超凡合并），**仅**上分帝 Web 英雄详情「走势」**无 STRATZ 时兜底**。STRATZ 周胜率/分路/`meta_board` 存 `web/data/stratz_hero_meta.json`（三卡与分路均取 **当前选中勋章 · 最近 1 周** `winWeek`；近 8 周列表从新到旧；分路在 8 周列表上方），对位 Top 存 `web/data/stratz_matchup_tops.json`（英雄详情「走势」与顶层「走势」榜 `#/trends` 优先用 STRATZ；「对位」Tab 只读），**禁止**合并进 relations/heroes，禁止 `pc/recommend.py` 读取。同英雄近期比赛出装/加点单独存 `web/data/hero_matches.json`，**仅** Web「近期比赛」Tab，禁止进 recommend。Valve Immortal 四区榜单独存 `web/data/leaderboards.json`，**仅** Web「排行」页选手榜，禁止进 recommend。主播目录单独存 `web/data/streamers.json`（手工名单 + 抖音 profile 补全；平台含抖音 / B 站 / 斗鱼），**仅** Web 顶层「主播」与英雄详情「主播」Tab，禁止进 recommend；由 `refresh_web` daily 软失败刷新粉丝等字段（不阻断整档）。
+- **克制 / 搭档数据**：机制克制/搭档只用定性边存 `shared/data/relations.json`（克制有向 + 搭档无向 + 理由），**不要**用胜率/场次表达机制克制，也**不要**写入 `shared/data/heroes.json`。Web 英雄页三视图：克制 / 被克制 / 搭档。OpenDota 段位胜率/场次单独存 `web/data/hero_stats.json`（近约 7 天窗；上场率 = 出场/(Σ出场/10)；冠绝样本过小时与超凡合并），**仅**上分帝 Web 英雄详情「走势」**无 STRATZ 时兜底**。STRATZ 周胜率/分路/`meta_board` 存 `web/data/stratz_hero_meta.json`（三卡与分路均取 **当前选中勋章 · 最近 1 周** `winWeek`；近 8 周列表从新到旧；分路在 8 周列表上方），对位 Top 存 `web/data/stratz_matchup_tops.json`（英雄详情「走势」与顶层「走势」榜 `/trends` 优先用 STRATZ；「对位」Tab 只读），**禁止**合并进 relations/heroes，禁止 `pc/recommend.py` 读取。同英雄近期比赛出装/加点单独存 `web/data/hero_matches.json`，**仅** Web「近期比赛」Tab，禁止进 recommend。Valve Immortal 四区榜单独存 `web/data/leaderboards.json`，**仅** Web「排行」页选手榜，禁止进 recommend。主播目录单独存 `web/data/streamers.json`（手工名单 + 抖音 profile 补全；平台含抖音 / B 站 / 斗鱼），**仅** Web 顶层「主播」与英雄详情「主播」Tab，禁止进 recommend；由 `refresh_web` daily 软失败刷新粉丝等字段（不阻断整档）。
 - **装备机制 / 怕的装备**：标签与技能汇总来自 Valve/OpenDota 自动抽取 + `item_tag_overrides.json`；`item_fears.py` 规则映射到英雄弱点。`fetch_item_counter_stats.py` 的对阵购买率/胜率差仅作为观测证据：须减去同装备全局基线，对全部英雄的已有机制候选只允许小幅调序；统计发现的新组合即使同时满足 `games≥100`、购买率提升 `≥3pp`、条件胜率差 `≥1.5pp`、两项双比例检验 `z≥1.96`，仍须确认机制成立后手工写入 overrides，禁止直接把高相关当因果克制。大段技能文案只进 `web/data/hero_abilities.json` / `items_meta.json`，**不要**塞进 `heroes.json`。本阶段仅上分帝 Web 展示，**不对局内出装推荐**。核心装「使用率」为 `hero_items` 列表内相对热度归一化，非绝对出场率、无段位维度。技能/物品「施加」类 tags（机制查询页）与技能 `dispellable`（效果能否被驱散）严格区分；**禁止**写入 recommend / relations。
- **机制查询页**：顶层 `#/mechanics[/{effect}]`；侧栏弱/强驱散 + 核心控制；结果为施加该效果的技能与物品。数据边界：仅 Web。
+- **机制查询页**：顶层 `/mechanics[/{effect}]`；侧栏弱/强驱散 + 核心控制；结果为施加该效果的技能与物品。数据边界：仅 Web。
 - **模板策略**：只维护 CDN 层。皮肤问题用会话策略（选人可改判、决策 `allow_revise=False`、best 帧偏 HERO_SELECTION），不要为皮肤加模板库，也不要复活 real 双层库。
 - **顶栏时机**：皮肤在**全员选完后**才上顶栏；本机进决策时别人可能还在选——须保持 `strategy_tail` 视觉，禁止「一进 STRATEGY 就永久停读」。
 - **坐标**：一律相对坐标（相对屏幕高 / 相对中心），勿写死像素分辨率。
@@ -221,7 +222,7 @@ python pc/evaluate.py
 - 改匹配阈值或裁切时，用 `pc/evaluate.py` / 标注帧验证，并更新 `CHANGELOG.md` 与必要时的 `ARCHITECTURE.md`。
 - 改上分帝 Web 视觉（色板、字号、间距、圆角、组件态）时先对齐 `DESIGN.md` 令牌，再改 `web/frontend/style.css`；勿引入未入规范的硬编码尺度。
 - 改 GSI cfg 时同步核对 `pc/gsi_setup.py` 与 Dota `gamestate_integration` 目录。
- 改 Web 路由形态（URL 段 / query 参数 / 默认值）时同步 `web/frontend/router.js`（`parseHash` / `serializeHash`）与 `app.js` 的 `applyPatch` 校验；新增可路由状态维度时在两处都加，并在 `syncStateToUrl` 调用点（含搜索 debounce）接好。Hash 路由不命中后端，`web/serve_relations.py` 无需改；`web/export_relations_site.py` 的导出文件列表须含 `router.js`。
+- 改 Web 路由形态（URL 段 / query 参数 / 默认值）时同步 `web/frontend/router.js`（`parseHash` / `serializeHash`）与 `app.js` 的 `applyPatch` 校验；新增可路由状态维度时在两处都加，并在 `syncStateToUrl` 调用点（含搜索 debounce）接好。History 深度链接依赖 `web/frontend/_redirects`（Cloudflare）与 `web/serve_relations.py` 的 SPA fallback；预渲染路径变更时同步 `web/seo_prerender.py` 与导出文件列表（含 `router.js` / `_redirects` / `robots.txt`）。
 - 改 `shared/data/heroes.json` 结构时同步 `shared/grid.py`（依赖 `attr` / `name_loc`）；`roles` 供 `pc/overlay.py` 使用；`aliases` 为中文口语/俗称（勿与 `name_loc` 重复），重跑 `pc/fetch_cdn_templates.py` 会按 `key` 合并保留；`tags` 为中文定位（核心/辅助/…/幻象，由 `roles`+幻想系推导），上分帝 Web 筛选 + 局内 `draft_archetypes` 缺口/推进画像共用；基础属性/血蓝等数值字段供上分帝 Web 详情条，勿塞机制文案。
 - 发版上分帝 Web 时同步 bump `web/export_relations_site.py` 的 `SITE_VERSION`、`web/frontend/config.js` 的同名变量，以及 `index.html` 里 `style.css`/`mobile-gate.js`/`config.js`/`router.js`/`app.js` 的 `?v=` 缓存戳；并写 `CHANGELOG.md`。
 - 不要重新引入 real 模板双层库、`cdn_penalty`、`build_library.py`。
@@ -240,13 +240,14 @@ Dota 2 (-gamestateintegration)
 - **打分**：关系边为主；阵容应对 / 缺口小幅加分；敌方 tags 软加分仅排序。
 - **展示**：网格左上角全量相关格标「克」（青）/「搭」（琥珀）/「补」（紫灰）；叠加层横条显示阵容分析；原因进日志与 `results` JSON，不写在每个格子旁。
 - **上分帝 Web**：`web/serve_relations.py` 本地开发服务；仿选将网格查看克制/被克制/搭档；数据手改 `shared/data/relations.json` 或从 xlsx 导入。
-  另有顶级 **排行**、**走势**（`#/trends[/bracket]`，近 8 周高胜率/上场率榜）、**物品**、**版本** 页；目录见 `web/data/item_shop.json` / `web/data/patches.json` / `web/data/leaderboards.json`。
-  英雄详情子标签含 **走势**（见上；`#/heroes/<key>/trends`）、**对位**（克制/被克/搭档三列 STRATZ Top；`#/heroes/<key>/matchups`）、**近期比赛**（`#/heroes/<key>/matches`）。排行页为 Immortal 四区选手榜。
- **Web Hash 路由**：状态（顶层标签 / 选中英雄 + 详情子标签 / 选中物品 / 选中版本 /
-  标签筛选 / 搜索）双向同步进 URL（`#/heroes/axe/core?tags=核心&q=axe`、`#/heroes/axe/trends` 等）。选 Hash
-  而非 Path 路由：静态导出（GitHub Pages）无需 rewrite、`web/serve_relations.py` 零改动、
-  相对路径子站亦可用。写 URL 用 `history.pushState`/`replaceState`（静默，不触发
-  `hashchange`，无回环）；前进/后退靠 `hashchange` 监听重 parse + 校验 + render。
+  另有顶级 **排行**、**走势**（`/trends[/bracket]`，近 8 周高胜率/上场率榜）、**物品**、**版本** 页；目录见 `web/data/item_shop.json` / `web/data/patches.json` / `web/data/leaderboards.json`。
+  英雄详情子标签含 **走势**（见上；`/heroes/<key>/trends`）、**对位**（克制/被克/搭档三列 STRATZ Top；`/heroes/<key>/matchups`）、**近期比赛**（`/heroes/<key>/matches`）。排行页为 Immortal 四区选手榜。
+- **Web History 路由 + SEO 预渲染**：状态（顶层标签 / 选中英雄 + 详情子标签 / 选中物品 / 选中版本 /
+  标签筛选 / 搜索）双向同步进路径 URL（`/heroes/axe/core?tags=核心&q=axe`、`/heroes/axe/trends` 等）。
+  旧 `#/...` 书签在装载时 `replaceState` 迁到路径。写 URL 用 `history.pushState`/`replaceState`
+  （静默，不触发 `popstate`）；前进/后退靠 `popstate`。Cloudflare `_redirects` 与
+  `serve_relations.py` 对未知深度路径回退 `index.html`。导出（`seo_prerender.py`）为首页、
+  顶层页、全英雄与机制效果写可抓取 HTML + `sitemap.xml` / `llms.txt`。
  搜索 debounce 300ms + `replaceState` 防刷历史栈；坏链接丢弃该项不崩。路由逻辑集中在
  `web/frontend/router.js`，`app.js` 只在 `main()` 装 `installRouter` + 各 state 变更点
  调 `syncStateToUrl`。
@@ -4,6 +4,13 @@

 ## [Unreleased]

+## [0.5.115] - 2026-07-30
+
+### Added
+
+- 上分帝 Web SEO / GEO：History 路径路由（兼容旧 `#/...` 书签）、首页 meta/OG/JSON-LD、路由态动态 `title`/`description`、爬虫跳过移动端门禁。
+- 导出预渲染：各英雄 `/heroes/{key}`、机制 `/mechanics/{effect}` 与顶层页可抓取 HTML；生成 `sitemap.xml` / `robots.txt` / `llms.txt`；Cloudflare `_redirects` SPA fallback；本地 `serve_relations` 同步支持深度链接刷新。
+
 ## [0.5.114] - 2026-07-30

 ### Added
@@ -56,7 +56,7 @@ python web/fetch_streamer_live.py  # 探测真实在播状态 → is_live/live_p
 python web/serve_relations.py       # 本地开发服务 http://127.0.0.1:8765（改 JSON 后刷新）
 ```

-Web 站点带 **Hash 路由**：`#/heroes/axe/core`、`#/heroes/axe/trends`、`#/heroes/axe/matchups`、`#/heroes/axe/matches`、`#/trends/legend`、`#/rankings`、`#/matches`（明星比赛，可 `#/matches/898754153`、`#/matches?origin=china`、`#/matches?page=2`）、`#/items/black_king_bar`、`#/patches/7.41`（`#/patches` = 最新，`#/rankings` = 中国区）；英雄页的定位标签筛选与搜索框也进 URL（`?tags=核心&q=axe`）。刷新保状态、可分享 / 深度链接、浏览器前进后退还原。`python web/export_relations_site.py` 导出的静态站点同样支持深度链接（可部署 GitHub Pages 或 Cloudflare Pages）。
+Web 站点带 **History 路径路由**：`/heroes/axe/core`、`/heroes/axe/trends`、`/heroes/axe/matchups`、`/heroes/axe/matches`、`/trends/legend`、`/rankings`、`/matches`（明星比赛，可 `/matches/898754153`、`/matches?origin=china`、`/matches?page=2`）、`/items/black_king_bar`、`/patches/7.41`（`/patches` = 最新，`/rankings` = 中国区）；英雄页的定位标签筛选与搜索框也进 URL（`?tags=核心&q=axe`）。旧 `#/...` 书签会自动迁移到路径。刷新保状态、可分享 / 深度链接、浏览器前进后退还原。导出时预渲染英雄/机制等页并生成 `sitemap.xml`（SEO）；Cloudflare Pages 用 `_redirects` 做 SPA fallback。

 ### 部署到 Cloudflare Pages

@@ -118,8 +118,8 @@ python web/refresh_web.py --tier patch --skip-deploy --skip-oss

 顶部另有 **排行** 页：Valve Immortal 四区榜各 Top 100（中国 / 欧洲 / 美洲 / 东南亚）；数据来自 `web/data/leaderboards.json`（`python web/fetch_leaderboards.py`）。
 顶部另有 **主播** 页：手工收录直播间/主页（抖音 + B 站）；列表先按是否在播、再按粉丝数降序；抖音式卡片布局为头像行（头像 \| 昵称 + 抖音号/获赞/粉丝或 B 站 UID/关注/粉丝 \|「关注」）+ 签名独立全宽行（最长 3 行）+ 常用英雄 + 可选精选高光视频（视口分档：远处不拉、近处 metadata、滚入中部 `canplay` 静音自动播放，同时仅 1 路全量缓冲；同名 JPG 封面作占位；有 `live_url` 时点头像进直播间；在播时粉环 +「直播」角标叠在环底、无间距——线上由访客触发的 `/api/live-status` 边缘函数刷新（缓存 5 分钟合并请求），本地 `serve_relations.py` 同步真实探测（内存缓存 60s）；角标只信本轮成功探测，`stale`/失败不沿用旧「直播中」；`data.json` 内 daily 探测值仅作接口返回前的首屏兜底）；数据来自 `web/data/streamers.json`（`python web/fetch_streamers.py` 补全；亦由 `web/refresh_web --tier daily` 每日软失败刷新）。视频与封面放 `web/assets/streamer_videos/`（gitignore），本地经 `/streamer-video/` 提供，部署前用 `web/_oss_static_assets.py upload` 同步至 OSS `streamer-video/`。
-顶部另有 **比赛** 页（`#/matches[/account_id][?origin=pro|china][&page=N]`，默认全部类型、每页 20 场）：侧栏可筛职业/国服与选手；明星选手 watchlist 近期联赛/锦标赛与天梯对局（出装/加点；卡片标职业/天梯/国服）；名单见 `web/data/pro_player_watchlist.json`，数据由 `python web/fetch_pro_matches.py --include-pubs` 拉取（亦进 daily）。
-顶部另有 **走势** 页：近 8 周各段位高胜率 / 上场率榜（`#/trends[/bracket]`；STRATZ 周胜率数据）。
+顶部另有 **比赛** 页（`/matches[/account_id][?origin=pro|china][&page=N]`，默认全部类型、每页 20 场）：侧栏可筛职业/国服与选手；明星选手 watchlist 近期联赛/锦标赛与天梯对局（出装/加点；卡片标职业/天梯/国服）；名单见 `web/data/pro_player_watchlist.json`，数据由 `python web/fetch_pro_matches.py --include-pubs` 拉取（亦进 daily）。
+顶部另有 **走势** 页：近 8 周各段位高胜率 / 上场率榜（`/trends[/bracket]`；STRATZ 周胜率数据）。
 顶部另有 **物品** 页：对齐 [官网商店物品](https://www.dota2.com.cn/items/index.htm) 的 **11 列竖排**（基础分类 5 列 + 合成分类 6 列），点选后详情显示在右侧（含合成组件）。目录来自 `web/data/item_shop.json`（`python web/fetch_item_shop.py`）。
 顶部另有 **版本** 页：默认展示最新版本完整改动（综合 / 物品 / 中立物品 / 英雄），下拉切换近一年其它版本；数据来自 `web/data/patches.json`（`python web/fetch_patches.py`，近一年窗口可用 `--days` / `--since` 调）。
 数据在 `shared/data/relations.json`（英雄关系，不用胜率）；选将全网格「克/搭/补」推荐读同一份文件，并用 `pc/draft_archetypes.py` 规则识别推进/全球流与敌我缺口（不接 AI）。热门装备来自 OpenDota 统计。**走势**优先 STRATZ（`web/data/stratz_hero_meta.json`，按勋章近 8 周 + 最近 1 周分路；`python web/fetch_stratz_meta.py`，需 token），OpenDota 各段位胜率/上场率/场次在 `web/data/hero_stats.json`（近约 7 天兜底；不进推荐）。**对位**数值 Top 仅用 `web/data/stratz_matchup_tops.json`（STRATZ 全局聚合相对优势，非走势页段位/周口径；与网格定性克/搭分开展示）。被克装备以描述标签 + 规则为主；`web/fetch_item_counter_stats.py` 另算敌方终局装备相对同装备全局基线的购买率提升/胜率差，对全部英雄的已有候选小幅调序；装备卡有数据时显示对阵该英雄的敌方队伍终局装备出现率。统计显著的新组合仍须确认机制成立后写入 `web/data/hero_fear_overrides.json`，避免“高相关但不克制”的误报。可改 `web/data/item_tag_overrides.json` 后重跑 `web/fetch_items_meta.py` 与 `web/item_fears.py`。**尚无对局内出装推荐**。
@@ -2,12 +2,15 @@

 Usage:
    python export_relations_site.py [--out dist/relations] [--with-videos]
-        [--ability-video-base URL] [--static-asset-base URL]
+        [--ability-video-base URL] [--static-asset-base URL] [--site-origin URL]

-Copies web/relations/ + a snapshot of the /api/data payload (data.json) +
+Copies web/frontend/ + a snapshot of the /api/data payload (data.json) +
 the referenced image assets into one directory, ready for any static host
 (GitHub Pages, Cloudflare Pages, nginx, ...).

+Also runs seo_prerender: crawlable HTML for heroes/mechanics/top pages,
+plus sitemap.xml / llms.txt / robots.txt / _redirects (History SPA).
+
 Notes:
 - Only already-cached assets are exported. For full ability-icon coverage
  run `python fetch_hero_abilities.py --icons-only` first.
@@ -50,9 +53,10 @@ from shared.paths import (
    WEB_DIST,
 )

+from seo_prerender import DEFAULT_SITE_ORIGIN, write_seo_bundle
 from serve_relations import WEB_DIR, build_payload

-SITE_VERSION = "0.5.114"
+SITE_VERSION = "0.5.115"
 DEFAULT_OSS_BASE = "https://climperor.oss-cn-shanghai.aliyuncs.com"


@@ -133,14 +137,17 @@ def write_config_js(
    ability_video_base: str,
    static_asset_base: str,
    site_version: str,
+    site_origin: str,
 ) -> None:
    """Write config.js consumed by app.js."""
    video = (ability_video_base or "").strip().rstrip("/")
    static = (static_asset_base or "").strip().rstrip("/")
    ver = (site_version or "").strip()
+    origin = (site_origin or "").strip().rstrip("/")
    (out / "config.js").write_text(
        "/* generated by export_relations_site.py — do not edit */\n"
        f"var SITE_VERSION = {json.dumps(ver, ensure_ascii=False)};\n"
+        f"var SITE_ORIGIN = {json.dumps(origin, ensure_ascii=False)};\n"
        f"var ABILITY_VIDEO_BASE = {json.dumps(video, ensure_ascii=False)};\n"
        f"var STATIC_ASSET_BASE = {json.dumps(static, ensure_ascii=False)};\n",
        encoding="utf-8",
@@ -170,6 +177,12 @@ def main() -> None:
        "when set, image dirs are not copied into dist; "
        "falls back to STATIC_ASSET_BASE env, else empty",
    )
+    ap.add_argument(
+        "--site-origin",
+        default=None,
+        help="canonical site origin for SEO (sitemap / og / config SITE_ORIGIN); "
+        "falls back to SITE_ORIGIN env, else https://dota2.refining.dev",
+    )
    args = ap.parse_args()

    video_base = (
@@ -182,6 +195,11 @@ def main() -> None:
        if args.static_asset_base is not None
        else os.environ.get("STATIC_ASSET_BASE", "")
    )
+    site_origin = (
+        args.site_origin
+        if args.site_origin is not None
+        else os.environ.get("SITE_ORIGIN", DEFAULT_SITE_ORIGIN)
+    )

    out = Path(args.out).resolve()
    if out == ROOT.resolve() or out.parent == out:
@@ -190,7 +208,16 @@ def main() -> None:
        shutil.rmtree(out)
    out.mkdir(parents=True)

-    for name in ("index.html", "router.js", "app.js", "style.css", "mobile-gate.js", "_headers"):
+    for name in (
+        "index.html",
+        "router.js",
+        "app.js",
+        "style.css",
+        "mobile-gate.js",
+        "_headers",
+        "_redirects",
+        "robots.txt",
+    ):
        src = WEB_DIR / name
        if src.is_file():
            shutil.copy2(src, out / name)
@@ -205,6 +232,7 @@ def main() -> None:
        ability_video_base=video_base,
        static_asset_base=static_base,
        site_version=SITE_VERSION,
+        site_origin=site_origin,
    )

    payload = build_payload()
@@ -217,6 +245,14 @@ def main() -> None:
        separators=(",", ":"),
    )

+    template_html = (WEB_DIR / "index.html").read_text(encoding="utf-8")
+    seo_counts = write_seo_bundle(
+        out,
+        template_html,
+        payload,
+        site_origin=site_origin,
+    )
+
    if static_base:
        counts = {
            name: 0
@@ -244,12 +280,17 @@ def main() -> None:
    print(f"exported static site -> {out}")
    for name, n in counts.items():
        print(f"  {name}/: {n} files")
+    print(
+        f"  seo prerender: top={seo_counts['top']} "
+        f"heroes={seo_counts['heroes']} mechanics={seo_counts['mechanics']} "
+        f"(+ sitemap.xml / llms.txt / robots.txt)"
+    )
    if n_functions:
        print(f"  functions/: {n_functions} files (Pages Functions)")
    if args.with_videos:
        print(f"  ability-video/: {n_videos} files")
    print(
-        f"  config.js SITE_VERSION={SITE_VERSION!r} "
+        f"  config.js SITE_VERSION={SITE_VERSION!r} SITE_ORIGIN={site_origin!r} "
        f"ABILITY_VIDEO_BASE={video_base!r} STATIC_ASSET_BASE={static_base!r}"
    )
    print(f"  total: {total / 1e6:.1f} MB")
@@ -0,0 +1,3 @@
+# Cloudflare Pages: static files win; these cover History deep links without a file.
+/heroes/:key/:tab  /heroes/:key/index.html  200
+/*  /index.html  200
@@ -1,4 +1,4 @@
-/* global fetch, document, ABILITY_VIDEO_BASE, STATIC_ASSET_BASE, SITE_VERSION */
+/* global fetch, document, ABILITY_VIDEO_BASE, STATIC_ASSET_BASE, SITE_VERSION, SITE_ORIGIN */

 function trimBase(raw) {
  return typeof raw === "string" ? raw.trim().replace(/\/+$/, "") : "";
@@ -76,6 +76,138 @@ function siteVersionLabel() {
  return v ? `v${v}` : "";
 }

+function siteOrigin() {
+  const configured = typeof SITE_ORIGIN === "string" ? SITE_ORIGIN.trim().replace(/\/+$/, "") : "";
+  if (configured) return configured;
+  if (typeof location !== "undefined" && location.origin) return location.origin;
+  return "";
+}
+
+const PAGE_SEO_LABELS = {
+  heroes: "英雄克制与搭档",
+  rankings: "Immortal 排行",
+  streamers: "主播",
+  matches: "明星比赛",
+  trends: "近 8 周走势",
+  mechanics: "机制查询",
+  items: "物品商店",
+  patches: "版本更新",
+};
+
+const DETAIL_TAB_SEO_LABELS = {
+  skills: "技能",
+  core: "核心装",
+  fears: "怕的装备",
+  trends: "走势",
+  matchups: "对位",
+  matches: "近期比赛",
+  streamers: "主播",
+  patches: "版本改动",
+};
+
+function setMetaByKey(attr, key, content) {
+  if (!content) return;
+  let el = document.querySelector(`meta[${attr}="${key}"]`);
+  if (!el) {
+    el = document.createElement("meta");
+    el.setAttribute(attr, key);
+    document.head.appendChild(el);
+  }
+  el.setAttribute("content", content);
+}
+
+function describeStateForSeo(st) {
+  const brand = "上分帝";
+  const page = st.page || "heroes";
+  let title = `${PAGE_SEO_LABELS[page] || "DOTA2"} — ${brand}`;
+  let description =
+    "Dota 2 英雄机制克制与搭档、段位走势、机制查询、物品与版本更新。";
+  let path = "/" + page;
+  try {
+    if (typeof serializeHash === "function") path = serializeHash(st) || "/heroes";
+  } catch (_) {
+    /* keep path */
+  }
+  if (!st.data) return { title, description, path };
+
+  if (page === "heroes" && st.selectedKey) {
+    const hero = heroByKey(st.selectedKey);
+    const name = (hero && hero.name_loc) || st.selectedKey;
+    const aliases = (hero && hero.aliases) || [];
+    const tab = DETAIL_TAB_SEO_LABELS[st.detailTab] || "详情";
+    title = `${name} ${tab} / 克制搭档 — ${brand}`;
+    const aliasBit = aliases.length ? `（${aliases.slice(0, 3).join("、")}）` : "";
+    description = `${name}${aliasBit}的 Dota 2 机制克制、被克制与搭档参考，以及技能、出装、走势与对位数据。`;
+  } else if (page === "mechanics") {
+    const mq = st.data.mechanic_query || {};
+    const labels = mq.labels || {};
+    const effect = st.mechanicEffect || "basic_dispel";
+    const label = labels[effect] || effect;
+    title = `${label} — 机制查询 — ${brand}`;
+    const blurb = (mq.blurbs && mq.blurbs[effect]) || "";
+    description = blurb || `查询 Dota 2 中施加「${label}」的技能与物品。`;
+  } else if (page === "items" && st.selectedItemKey) {
+    const meta = shopItem(st.selectedItemKey);
+    const name = (meta && (meta.name_loc || meta.dname)) || st.selectedItemKey;
+    title = `${name} — 物品 — ${brand}`;
+    description = `${name} 的合成、描述与机制标签（上分帝物品页）。`;
+  } else if (page === "patches") {
+    const ver =
+      st.selectedPatch ||
+      ((st.data.patches && st.data.patches[0]) || {}).version;
+    if (ver) {
+      title = `版本 ${ver} — ${brand}`;
+      description = `Dota 2 ${ver} 游戏性更新摘要（上分帝版本页）。`;
+    }
+  } else if (page === "trends") {
+    title = `近 8 周走势 — ${brand}`;
+    description = "各勋章段位近 8 周英雄胜率与上场率走势榜。";
+  } else if (page === "rankings") {
+    title = `Immortal 排行榜 — ${brand}`;
+    description = "Valve Immortal 四区 Top100 选手榜。";
+  } else if (page === "streamers") {
+    title = `Dota 2 主播 — ${brand}`;
+    description = "精选 Dota 2 主播目录与高光（抖音 / B 站 / 斗鱼）。";
+  } else if (page === "matches") {
+    title = `明星比赛 — ${brand}`;
+    description = "明星选手近期职业与国服对局、终局出装与加点。";
+  } else if (page === "heroes") {
+    title = `英雄克制与搭档 — ${brand}`;
+    description =
+      "按英雄浏览定性克制 / 被克制 / 搭档理由，以及技能、核心装、走势与对位。";
+  }
+
+  return { title, description, path };
+}
+
+function updateDocumentMeta() {
+  if (typeof document === "undefined" || !document.title) return;
+  const { title, description, path } = describeStateForSeo(state);
+  document.title = title;
+  setMetaByKey("name", "description", description);
+  setMetaByKey("property", "og:title", title);
+  setMetaByKey("property", "og:description", description);
+  setMetaByKey("name", "twitter:title", title);
+  setMetaByKey("name", "twitter:description", description);
+  const origin = siteOrigin();
+  if (origin) {
+    const url = origin + (path.startsWith("/") ? path : `/${path}`);
+    setMetaByKey("property", "og:url", url);
+    let link = document.querySelector('link[rel="canonical"]');
+    if (!link) {
+      link = document.createElement("link");
+      link.setAttribute("rel", "canonical");
+      document.head.appendChild(link);
+    }
+    link.setAttribute("href", url);
+  }
+}
+
+function clearSeoPrerender() {
+  const el = document.getElementById("seo-prerender");
+  if (el) el.remove();
+}
+
 function renderPatchesSiteVersion() {
  const el = $("#patches-site-version");
  if (!el) return;
@@ -5468,6 +5600,7 @@ function render() {
  } else {
    renderPatches();
  }
+  updateDocumentMeta();
 }

 // Search-box URL sync timers (debounced replace so typing does not spam history).
@@ -5698,7 +5831,8 @@ async function main() {
  if (brandLogo) brandLogo.src = assetUrl("/ui-icon/dota2_logo_wordmark.png");
  const favicon = document.querySelector('link[rel="icon"]');
  if (favicon) favicon.href = assetUrl("/ui-icon/dota2_logo.png");
-  const res = await fetch("data.json");
+  // Absolute path: History routes like /heroes/axe must not resolve data.json relatively.
+  const res = await fetch("/data.json");
  if (!res.ok) throw new Error(`data.json: HTTP ${res.status}`);
  state.data = await res.json();
  state.data.relations = state.data.relations || { counters: [], synergies: [] };
@@ -5755,6 +5889,7 @@ async function main() {
  _innateAbilityKeys = null;
  installRouter({ getState: () => state, applyPatch });
  applyUrlToState();
+  clearSeoPrerender();
  refreshStreamerLiveStatus();
 }

@@ -1,5 +1,6 @@
 /* Local defaults; production export overwrites via export_relations_site.py. */
-var SITE_VERSION = "0.5.114";
+var SITE_VERSION = "0.5.115";
+var SITE_ORIGIN = "";
 var ABILITY_VIDEO_BASE = "";
 var STATIC_ASSET_BASE = "";

@@ -3,13 +3,67 @@
 <head>
  <meta charset="UTF-8" />
  <meta name="viewport" content="width=device-width, initial-scale=1" />
-  <title>DOTA2 上分帝</title>
+  <title>DOTA2 上分帝 — 英雄克制 / 搭档 / 走势 / 机制</title>
+  <meta
+    name="description"
+    content="上分帝（Climperor）：Dota 2 英雄机制克制与搭档、段位走势、机制查询、物品与版本更新。定性关系理由 + 公开数据走势，助你上分。"
+  />
+  <meta name="keywords" content="DOTA2,上分帝,Climperor,英雄克制,搭档,胜率走势,机制查询,驱散,版本更新" />
+  <link rel="canonical" href="https://dota2.refining.dev/" />
+  <meta property="og:type" content="website" />
+  <meta property="og:site_name" content="上分帝" />
+  <meta property="og:locale" content="zh_CN" />
+  <meta property="og:title" content="DOTA2 上分帝 — 英雄克制 / 搭档 / 走势 / 机制" />
+  <meta
+    property="og:description"
+    content="Dota 2 英雄机制克制与搭档、段位走势、机制查询、物品与版本更新。"
+  />
+  <meta property="og:url" content="https://dota2.refining.dev/" />
+  <meta property="og:image" content="https://climperor.oss-cn-shanghai.aliyuncs.com/ui-icon/dota2_logo_wordmark.png" />
+  <meta name="twitter:card" content="summary" />
+  <meta name="twitter:title" content="DOTA2 上分帝" />
+  <meta
+    name="twitter:description"
+    content="Dota 2 英雄机制克制与搭档、段位走势、机制查询、物品与版本更新。"
+  />
+  <script type="application/ld+json" id="seo-jsonld">
+  {
+    "@context": "https://schema.org",
+    "@type": "WebSite",
+    "name": "上分帝",
+    "alternateName": ["Climperor", "DOTA2 上分帝"],
+    "url": "https://dota2.refining.dev/",
+    "inLanguage": "zh-CN",
+    "description": "Dota 2 英雄机制克制与搭档、段位走势、机制查询、物品与版本更新。",
+    "potentialAction": {
+      "@type": "SearchAction",
+      "target": "https://dota2.refining.dev/heroes?q={search_term_string}",
+      "query-input": "required name=search_term_string"
+    }
+  }
+  </script>
  <link rel="icon" href="/ui-icon/dota2_logo.png" type="image/png" />
-  <link rel="stylesheet" href="/style.css?v=0.5.114" />
-  <script src="/mobile-gate.js?v=0.5.114"></script>
+  <link rel="stylesheet" href="/style.css?v=0.5.115" />
+  <script src="/mobile-gate.js?v=0.5.115"></script>
 </head>
 <body>
  <h1 class="sr-only">DOTA2 上分帝</h1>
+  <aside id="seo-prerender" class="seo-prerender">
+    <p>
+      上分帝（Climperor）是面向 Dota 2 的选将与机制参考站：浏览英雄克制与搭档理由、
+      段位走势、机制查询（驱散与控制）、物品商店与版本更新。请使用桌面浏览器访问完整交互界面。
+    </p>
+    <ul>
+      <li><a href="/heroes">英雄克制与搭档</a></li>
+      <li><a href="/mechanics">机制查询</a></li>
+      <li><a href="/trends">近 8 周走势</a></li>
+      <li><a href="/items">物品商店</a></li>
+      <li><a href="/patches">版本更新</a></li>
+      <li><a href="/rankings">Immortal 排行</a></li>
+      <li><a href="/streamers">主播</a></li>
+      <li><a href="/matches">明星比赛</a></li>
+    </ul>
+  </aside>
  <div id="mobile-gate" class="mobile-gate" role="dialog" aria-labelledby="mobile-gate-title" aria-modal="true">
    <div class="mobile-gate-card">
      <div class="mobile-gate-brand" aria-hidden="true">
@@ -178,8 +232,8 @@

  <section class="detail" id="detail" aria-live="polite"></section>

-  <script src="/config.js?v=0.5.114"></script>
-  <script src="/router.js?v=0.5.114"></script>
-  <script src="/app.js?v=0.5.114"></script>
+  <script src="/config.js?v=0.5.115"></script>
+  <script src="/router.js?v=0.5.115"></script>
+  <script src="/app.js?v=0.5.115"></script>
 </body>
 </html>
@@ -2,8 +2,16 @@
 * Early mobile client gate (loaded from <head>).
 * Sets html.mobile-client before paint; wires the demand button after DOM ready.
 * Exposes window.__CLIMPEROR_MOBILE__ for app.js to skip the desktop boot path.
+ * Search / AI crawlers skip the gate so prerendered HTML stays indexable.
 */
 (function () {
+  function isCrawler() {
+    var ua = navigator.userAgent || "";
+    return /bot|crawl|spider|slurp|bingpreview|facebookexternalhit|embedly|quora link preview|pinterest|redditbot|linkedinbot|twitterbot|whatsapp|google-inspectiontool|bytespider|baiduspider|yandex|duckduckbot|applebot|semrush|ahrefs|gptbot|claudebot|anthropic|perplexity|oai-searchbot|chatgpt/i.test(
+      ua
+    );
+  }
+
  function isMobileClient() {
    var ua = navigator.userAgent || "";
    if (/Android|webOS|iPhone|iPad|iPod|BlackBerry|IEMobile|Opera Mini/i.test(ua)) {
@@ -16,6 +24,11 @@
    return false;
  }

+  if (isCrawler()) {
+    window.__CLIMPEROR_MOBILE__ = false;
+    return;
+  }
+
  var mobile = isMobileClient();
  window.__CLIMPEROR_MOBILE__ = mobile;
  if (!mobile) return;
@@ -0,0 +1,4 @@
+User-agent: *
+Allow: /
+
+Sitemap: https://dota2.refining.dev/sitemap.xml
@@ -1,38 +1,43 @@
 /* global window, history, location */

 /**
- * Hash-based router for the Climperor web site (web/relations).
+ * Path-based router for the Climperor web site (web/frontend).
 *
 * Synchronizes the browser URL with app state across these dimensions:
- *   - page: heroes | rankings | matches | streamers | trends | mechanics | items | patches (top-level tab)
+ *   - page: heroes | rankings | matches | streamers | trends | mechanics | items | patches
 *   - hero: selected hero key + detail sub-tab
 *     (skills|core|fears|trends|matchups|matches|streamers|patches; legacy stats → trends)
 *   - rankings: Immortal leaderboard region
- *     #/rankings[/region]  (legacy #/rankings/meta[/bracket] → players / china)
+ *     /rankings[/region]  (legacy /rankings/meta[/bracket] → players / china)
 *   - matches: star-player recent matches (pro_matches)
- *     #/matches[/account_id][?origin=pro|china][&page=N]
+ *     /matches[/account_id][?origin=pro|china][&page=N]
 *     (default origin all omitted; page=1 omitted)
- *   - streamers: curated Douyin streamer directory
- *     #/streamers
+ *   - streamers: curated streamer directory
+ *     /streamers
 *   - trends: medal bracket for the 8-week win/pick board
- *     #/trends[/bracket][?sort=pr]  (default bracket legend, sort wr omitted)
+ *     /trends[/bracket][?sort=pr]  (default bracket legend, sort wr omitted)
 *   - mechanics: applies-effect query (dispel / CC)
- *     #/mechanics[/{effect}]  (default effect basic_dispel omitted)
+ *     /mechanics[/{effect}]  (default effect basic_dispel omitted)
 *   - item: selected shop item key (items page)
 *   - patch: selected version string (patches page; latest when absent)
 *   - query params: tags=csv (heroes), q=search text (heroes + items)
 *
+ * Legacy hash URLs (#/heroes/...) are migrated once to path URLs via replaceState.
+ *
 * Two-way binding without feedback loops:
 *   - App calls syncStateToUrl() after each state mutation. We use
 *     history.pushState/replaceState, which update the URL silently
- *     (no hashchange event fires) — so applyUrlToState is never re-entered.
- *   - Back/forward navigation fires hashchange → applyUrlToState parses
- *     the new hash, hands a validated patch back to app, and app re-renders.
+ *     (no popstate) — so applyUrlToState is never re-entered.
+ *   - Back/forward navigation fires popstate → applyUrlToState parses
+ *     the new path, hands a validated patch back to app, and app re-renders.
+ *
+ * API names parseHash / serializeHash are kept for call-site stability; both
+ * now operate on pathname + search (not location.hash).
 *
 * Depends on app.js for state shape and render(); installed via installRouter().
 */

-const ROUTE_DEFAULT = "#/heroes";
+const ROUTE_DEFAULT = "/heroes";
 const VALID_PAGES = ["heroes", "rankings", "matches", "streamers", "trends", "mechanics", "items", "patches"];
 const VALID_DETAIL_TABS = [
  "skills",
@@ -85,11 +90,20 @@ let _deps = null;
 */
 function installRouter(deps) {
  _deps = deps;
-  window.addEventListener("hashchange", applyUrlToState);
+  migrateLegacyHashRoute();
+  window.addEventListener("popstate", applyUrlToState);
 }

-/** Parse a hash string into a state patch (all fields nullable). */
-function parseHash(hash) {
+/** One-shot: #/heroes/... → /heroes/... (bookmarks / shared links). */
+function migrateLegacyHashRoute() {
+  const hash = window.location.hash || "";
+  if (!hash.startsWith("#/")) return;
+  const next = hash.slice(1) || ROUTE_DEFAULT;
+  history.replaceState(null, "", next);
+}
+
+/** Parse a path (+ optional query) into a state patch (all fields nullable). */
+function parseHash(hashOrPath) {
  const out = {
    page: null,
    heroKey: null,
@@ -107,14 +121,17 @@ function parseHash(hash) {
    query: null,
    itemQuery: null,
  };
-  let raw = hash || "";
+  let raw = hashOrPath || "";
  if (raw.startsWith("#")) raw = raw.slice(1);
  if (!raw.startsWith("/")) raw = "/" + raw;
  const qIdx = raw.indexOf("?");
  const pathPart = qIdx >= 0 ? raw.slice(0, qIdx) : raw;
  const queryPart = qIdx >= 0 ? raw.slice(qIdx + 1) : "";
  const segs = pathPart.split("/").filter(Boolean);
-  if (!segs.length) return out;
+  if (!segs.length) {
+    out.page = "heroes";
+    return out;
+  }
  const page = segs[0];
  if (!VALID_PAGES.includes(page)) return out;
  out.page = page;
@@ -122,7 +139,7 @@ function parseHash(hash) {
    if (segs[1]) out.heroKey = safeDecode(segs[1]);
    if (segs[2]) out.detailTab = safeDecode(segs[2]);
  } else if (page === "rankings") {
-    // Legacy #/rankings/meta[/bracket] → Immortal players board (default region).
+    // Legacy /rankings/meta[/bracket] → Immortal players board (default region).
    if (segs[1] && segs[1] !== "meta") {
      out.rankingRegion = safeDecode(segs[1]);
    }
@@ -179,56 +196,56 @@ function safeDecode(s) {
  }
 }

-/** Serialize current app state into a hash string (with leading '#'). */
+/** Serialize current app state into a path string (leading '/', optional ?query). */
 function serializeHash(state) {
  if (!state || !VALID_PAGES.includes(state.page)) return ROUTE_DEFAULT;
-  let hash = "#/" + state.page;
+  let path = "/" + state.page;
  if (state.page === "heroes") {
    if (state.selectedKey) {
-      hash += "/" + encodeURIComponent(state.selectedKey);
+      path += "/" + encodeURIComponent(state.selectedKey);
      let tab = state.detailTab;
      if (tab === "stats") tab = "trends"; // normalize legacy alias in URL
      if (tab && VALID_DETAIL_TABS.includes(tab) && tab !== "stats") {
-        hash += "/" + encodeURIComponent(tab);
+        path += "/" + encodeURIComponent(tab);
      }
    }
  } else if (state.page === "rankings") {
-    // Omit region when it is the default (china) — bare #/rankings means China.
+    // Omit region when it is the default (china) — bare /rankings means China.
    const region = state.rankingRegion || DEFAULT_RANKING_REGION;
    if (region && region !== DEFAULT_RANKING_REGION) {
-      hash += "/" + encodeURIComponent(region);
+      path += "/" + encodeURIComponent(region);
    }
  } else if (state.page === "matches") {
    if (state.matchesPlayerId) {
-      hash += "/" + encodeURIComponent(String(state.matchesPlayerId));
+      path += "/" + encodeURIComponent(String(state.matchesPlayerId));
    }
  } else if (state.page === "trends") {
-    // Omit bracket when it is the default (legend) — bare #/trends means legend.
+    // Omit bracket when it is the default (legend) — bare /trends means legend.
    const bracket = state.trendsBracket || DEFAULT_TRENDS_BRACKET;
    if (bracket && bracket !== DEFAULT_TRENDS_BRACKET) {
-      hash += "/" + encodeURIComponent(bracket);
+      path += "/" + encodeURIComponent(bracket);
    }
    const sort = state.trendsSort || DEFAULT_TRENDS_SORT;
    if (VALID_TRENDS_SORTS.includes(sort) && sort !== DEFAULT_TRENDS_SORT) {
-      hash += "?sort=" + encodeURIComponent(TRENDS_SORT_URL[sort] || sort);
+      path += "?sort=" + encodeURIComponent(TRENDS_SORT_URL[sort] || sort);
    }
  } else if (state.page === "mechanics") {
    const effect = state.mechanicEffect || DEFAULT_MECHANIC_EFFECT;
    if (effect && effect !== DEFAULT_MECHANIC_EFFECT) {
-      hash += "/" + encodeURIComponent(effect);
+      path += "/" + encodeURIComponent(effect);
    }
  } else if (state.page === "items") {
    if (state.selectedItemKey) {
-      hash += "/" + encodeURIComponent(state.selectedItemKey);
+      path += "/" + encodeURIComponent(state.selectedItemKey);
    }
  } else if (state.page === "patches") {
-    // Omit version when it is the latest — bare #/patches means "latest".
+    // Omit version when it is the latest — bare /patches means "latest".
    if (state.selectedPatch) {
      const patches = (state.data && state.data.patches) || [];
      const isLatest =
        patches.length > 0 && patches[0].version === state.selectedPatch;
      if (!isLatest) {
-        hash += "/" + encodeURIComponent(state.selectedPatch);
+        path += "/" + encodeURIComponent(state.selectedPatch);
      }
    }
  }
@@ -239,12 +256,12 @@ function serializeHash(state) {
    }
    if (state.query) qs.set("q", state.query);
    const s = qs.toString();
-    if (s) hash += "?" + s;
+    if (s) path += "?" + s;
  } else if (state.page === "items") {
    const qs = new URLSearchParams();
    if (state.itemQuery) qs.set("q", state.itemQuery);
    const s = qs.toString();
-    if (s) hash += "?" + s;
+    if (s) path += "?" + s;
  } else if (state.page === "matches") {
    const qs = new URLSearchParams();
    const origin = state.matchesOrigin || DEFAULT_MATCHES_ORIGIN;
@@ -254,29 +271,33 @@ function serializeHash(state) {
    const pageNum = Number(state.matchesPage) || 1;
    if (pageNum > 1) qs.set("page", String(pageNum));
    const s = qs.toString();
-    if (s) hash += "?" + s;
+    if (s) path += "?" + s;
  }
-  return hash;
+  return path;
+}
+
+function currentRoutePath() {
+  return (window.location.pathname || "/") + (window.location.search || "");
 }

 /** Write current state into the URL. Use replace:true for high-frequency
 *  updates (search typing, tag toggles) to avoid history-stack spam. */
 function syncStateToUrl({ replace = false } = {}) {
  if (!_deps) return;
-  const hash = serializeHash(_deps.getState());
-  if (hash === window.location.hash) return; // no-op when state matches URL
+  const path = serializeHash(_deps.getState());
+  if (path === currentRoutePath()) return; // no-op when state matches URL
  if (replace) {
-    history.replaceState(null, "", hash);
+    history.replaceState(null, "", path);
  } else {
-    history.pushState(null, "", hash);
+    history.pushState(null, "", path);
  }
 }

-/** Parse current location.hash, hand it to app for validation+merge, render.
- *  Invoked on hashchange (back/forward / direct location.hash writes) and
- *  once at startup for deep-link support. */
+/** Parse current location path+search, hand it to app for validation+merge.
+ *  Invoked on popstate (back/forward) and once at startup for deep-link support. */
 function applyUrlToState() {
  if (!_deps) return;
-  const patch = parseHash(window.location.hash);
+  const path = currentRoutePath();
+  const patch = parseHash(path === "/" ? "/heroes" : path);
  _deps.applyPatch(patch);
 }
@@ -135,7 +135,8 @@ body {
  color: var(--muted);
  line-height: 1.4;
 }
-.sr-only {
+.sr-only,
+.seo-prerender {
  position: absolute;
  width: 1px;
  height: 1px;
@@ -65,6 +65,8 @@ INPUT_WATCH = [
    WEB_FRONTEND / "mobile-gate.js",
    WEB_FRONTEND / "config.js",
    WEB_FRONTEND / "_headers",
+    WEB_FRONTEND / "_redirects",
+    WEB_FRONTEND / "robots.txt",
 ]
 FRONTEND_FUNCTIONS_DIR = WEB_FRONTEND / "functions"
 ASSET_DIRS = [
@@ -0,0 +1,455 @@
+"""Build SEO prerender HTML, sitemap, and llms.txt for Climperor Web export.
+
+Writes crawlable path pages under dist (heroes / mechanics / top-level tabs)
+while keeping the same SPA shell for hydration.
+"""
+
+from __future__ import annotations
+
+import html
+import json
+import re
+from pathlib import Path
+from xml.sax.saxutils import escape as xml_escape
+
+DEFAULT_SITE_ORIGIN = "https://dota2.refining.dev"
+
+TOP_PAGES: list[tuple[str, str, str]] = [
+    ("/", "英雄克制与搭档", "按英雄浏览定性克制、被克制与搭档理由，以及技能与走势。"),
+    ("/heroes", "英雄克制与搭档", "Dota 2 英雄机制克制与搭档目录。"),
+    ("/mechanics", "机制查询", "查询施加驱散与控制等效果的技能与物品。"),
+    ("/trends", "近 8 周走势", "各勋章段位近 8 周英雄胜率与上场率走势榜。"),
+    ("/items", "物品商店", "基础与合成分类物品目录。"),
+    ("/patches", "版本更新", "近一年游戏性更新摘要。"),
+    ("/rankings", "Immortal 排行", "Valve Immortal 四区 Top100。"),
+    ("/streamers", "主播", "精选 Dota 2 主播目录。"),
+    ("/matches", "明星比赛", "明星选手近期职业与国服对局。"),
+]
+
+_TITLE_RE = re.compile(r"<title>[^<]*</title>", re.I)
+_DESC_RE = re.compile(
+    r'<meta\s+name="description"\s+content="[^"]*"\s*/?>',
+    re.I,
+)
+_CANONICAL_RE = re.compile(
+    r'<link\s+rel="canonical"\s+href="[^"]*"\s*/?>',
+    re.I,
+)
+_OG_TITLE_RE = re.compile(
+    r'<meta\s+property="og:title"\s+content="[^"]*"\s*/?>',
+    re.I,
+)
+_OG_DESC_RE = re.compile(
+    r'<meta\s+property="og:description"\s+content="[^"]*"\s*/?>',
+    re.I,
+)
+_OG_URL_RE = re.compile(
+    r'<meta\s+property="og:url"\s+content="[^"]*"\s*/?>',
+    re.I,
+)
+_TW_TITLE_RE = re.compile(
+    r'<meta\s+name="twitter:title"\s+content="[^"]*"\s*/?>',
+    re.I,
+)
+_TW_DESC_RE = re.compile(
+    r'<meta\s+name="twitter:description"\s+content="[^"]*"\s*/?>',
+    re.I,
+)
+_JSONLD_RE = re.compile(
+    r'<script type="application/ld\+json" id="seo-jsonld">.*?</script>',
+    re.I | re.S,
+)
+_SEO_ASIDE_RE = re.compile(
+    r'<aside id="seo-prerender" class="seo-prerender">.*?</aside>',
+    re.I | re.S,
+)
+
+
+def _esc(s: object) -> str:
+    return html.escape("" if s is None else str(s), quote=True)
+
+
+def _abs(origin: str, path: str) -> str:
+    base = (origin or DEFAULT_SITE_ORIGIN).rstrip("/")
+    if not path.startswith("/"):
+        path = "/" + path
+    return base + path
+
+
+def _hero_name_map(payload: dict) -> dict[str, str]:
+    out: dict[str, str] = {}
+    for h in payload.get("heroes") or []:
+        if not isinstance(h, dict):
+            continue
+        key = h.get("key")
+        if not key:
+            continue
+        out[str(key)] = str(h.get("name_loc") or key)
+    return out
+
+
+def _relation_lists(payload: dict, hero_key: str) -> tuple[list[str], list[str], list[str]]:
+    names = _hero_name_map(payload)
+    rel = payload.get("relations") or {}
+    counters_out: list[str] = []
+    countered_out: list[str] = []
+    syn_out: list[str] = []
+    for edge in rel.get("counters") or []:
+        if not isinstance(edge, dict):
+            continue
+        a, b = edge.get("a"), edge.get("b")
+        reason = (edge.get("reason") or "").strip()
+        if a == hero_key and b:
+            label = names.get(str(b), str(b))
+            counters_out.append(f"{label}" + (f"：{reason}" if reason else ""))
+        elif b == hero_key and a:
+            label = names.get(str(a), str(a))
+            countered_out.append(f"{label}" + (f"：{reason}" if reason else ""))
+    for edge in rel.get("synergies") or []:
+        if not isinstance(edge, dict):
+            continue
+        a, b = edge.get("a"), edge.get("b")
+        reason = (edge.get("reason") or "").strip()
+        peer = None
+        if a == hero_key and b:
+            peer = str(b)
+        elif b == hero_key and a:
+            peer = str(a)
+        if peer:
+            label = names.get(peer, peer)
+            syn_out.append(f"{label}" + (f"：{reason}" if reason else ""))
+    return counters_out[:12], countered_out[:12], syn_out[:12]
+
+
+def _ul(items: list[str]) -> str:
+    if not items:
+        return "<p>暂无条目</p>"
+    lis = "".join(f"<li>{_esc(x)}</li>" for x in items)
+    return f"<ul>{lis}</ul>"
+
+
+def _hero_seo_body(hero: dict, payload: dict) -> str:
+    key = str(hero.get("key") or "")
+    name = str(hero.get("name_loc") or key)
+    aliases = [str(a) for a in (hero.get("aliases") or []) if a]
+    tags = [str(t) for t in (hero.get("tags") or []) if t]
+    counters, countered, syns = _relation_lists(payload, key)
+    alias_bit = f"（别名：{'、'.join(_esc(a) for a in aliases)}）" if aliases else ""
+    tag_bit = f"<p>定位：{'、'.join(_esc(t) for t in tags)}</p>" if tags else ""
+    return (
+        f"<article>"
+        f"<h1>{_esc(name)} — 克制与搭档</h1>"
+        f"<p>{_esc(name)}{alias_bit}的 Dota 2 机制克制、被克制与搭档参考（上分帝定性关系，非胜率因果）。</p>"
+        f"{tag_bit}"
+        f"<h2>克制</h2>{_ul(counters)}"
+        f"<h2>被克制</h2>{_ul(countered)}"
+        f"<h2>搭档</h2>{_ul(syns)}"
+        f"<p><a href=\"/heroes\">返回英雄目录</a> · "
+        f"<a href=\"/mechanics\">机制查询</a></p>"
+        f"</article>"
+    )
+
+
+def _mechanic_seo_body(effect: str, payload: dict) -> str:
+    mq = payload.get("mechanic_query") or {}
+    labels = mq.get("labels") or {}
+    blurbs = mq.get("blurbs") or {}
+    label = labels.get(effect) or effect
+    blurb = blurbs.get(effect) or f"列出施加「{label}」的技能与物品。"
+    names = _hero_name_map(payload)
+    abil_lines: list[str] = []
+    by_hero = ((payload.get("hero_abilities") or {}).get("by_hero")) or {}
+    for hkey, cell in by_hero.items():
+        if not isinstance(cell, dict):
+            continue
+        hname = names.get(str(hkey), str(hkey))
+        for ab in cell.get("abilities") or []:
+            if not isinstance(ab, dict):
+                continue
+            if effect not in (ab.get("tags") or []):
+                continue
+            aname = ab.get("name_loc") or ab.get("key") or ""
+            abil_lines.append(f"{hname} · {aname}")
+    abil_lines = sorted(set(abil_lines), key=lambda s: s)[:80]
+    item_lines: list[str] = []
+    for row in (payload.get("items_meta") or {}).values():
+        if not isinstance(row, dict):
+            continue
+        if effect not in (row.get("tags") or []):
+            continue
+        item_lines.append(str(row.get("name_loc") or row.get("key") or ""))
+    item_lines = sorted({x for x in item_lines if x})[:40]
+    return (
+        f"<article>"
+        f"<h1>{_esc(label)} — 机制查询</h1>"
+        f"<p>{_esc(blurb)}</p>"
+        f"<h2>技能（{len(abil_lines)}）</h2>{_ul(abil_lines)}"
+        f"<h2>物品（{len(item_lines)}）</h2>{_ul(item_lines)}"
+        f"<p><a href=\"/mechanics\">全部机制</a> · <a href=\"/heroes\">英雄</a></p>"
+        f"</article>"
+    )
+
+
+def _top_seo_body(path: str, title: str, description: str, payload: dict) -> str:
+    hero_links = []
+    for h in (payload.get("heroes") or [])[:40]:
+        if not isinstance(h, dict) or not h.get("key"):
+            continue
+        key = str(h["key"])
+        name = str(h.get("name_loc") or key)
+        hero_links.append(f'<li><a href="/heroes/{_esc(key)}">{_esc(name)}</a></li>')
+    mq = payload.get("mechanic_query") or {}
+    labels = mq.get("labels") or {}
+    mech_links = []
+    for effect in mq.get("order") or []:
+        label = labels.get(effect) or effect
+        href = "/mechanics" if effect == "basic_dispel" else f"/mechanics/{effect}"
+        mech_links.append(f'<li><a href="{_esc(href)}">{_esc(label)}</a></li>')
+    extra = ""
+    if path in ("/", "/heroes"):
+        extra = f"<h2>英雄目录（部分）</h2><ul>{''.join(hero_links)}</ul>"
+    if path in ("/", "/mechanics"):
+        extra += f"<h2>机制效果</h2><ul>{''.join(mech_links)}</ul>"
+    return (
+        f"<article>"
+        f"<h1>{_esc(title)} — 上分帝</h1>"
+        f"<p>{_esc(description)}</p>"
+        f"{extra}"
+        f"</article>"
+    )
+
+
+def _jsonld_website(origin: str, title: str, description: str, url: str) -> str:
+    payload = {
+        "@context": "https://schema.org",
+        "@graph": [
+            {
+                "@type": "WebSite",
+                "name": "上分帝",
+                "alternateName": ["Climperor", "DOTA2 上分帝"],
+                "url": origin.rstrip("/") + "/",
+                "inLanguage": "zh-CN",
+                "description": "Dota 2 英雄机制克制与搭档、段位走势、机制查询、物品与版本更新。",
+                "potentialAction": {
+                    "@type": "SearchAction",
+                    "target": origin.rstrip("/") + "/heroes?q={search_term_string}",
+                    "query-input": "required name=search_term_string",
+                },
+            },
+            {
+                "@type": "WebPage",
+                "name": title,
+                "description": description,
+                "url": url,
+                "isPartOf": {"@type": "WebSite", "name": "上分帝", "url": origin.rstrip("/") + "/"},
+                "inLanguage": "zh-CN",
+            },
+        ],
+    }
+    body = json.dumps(payload, ensure_ascii=False, indent=2)
+    return f'<script type="application/ld+json" id="seo-jsonld">\n{body}\n  </script>'
+
+
+def inject_seo(
+    template: str,
+    *,
+    title: str,
+    description: str,
+    canonical: str,
+    seo_body_html: str,
+    origin: str,
+) -> str:
+    """Replace head SEO tags and #seo-prerender body in the SPA shell."""
+    out = template
+    out = _TITLE_RE.sub(f"<title>{_esc(title)}</title>", out, count=1)
+    out = _DESC_RE.sub(
+        f'<meta\n    name="description"\n    content="{_esc(description)}"\n  />',
+        out,
+        count=1,
+    )
+    out = _CANONICAL_RE.sub(
+        f'<link rel="canonical" href="{_esc(canonical)}" />',
+        out,
+        count=1,
+    )
+    out = _OG_TITLE_RE.sub(
+        f'<meta property="og:title" content="{_esc(title)}" />',
+        out,
+        count=1,
+    )
+    out = _OG_DESC_RE.sub(
+        f'<meta\n    property="og:description"\n    content="{_esc(description)}"\n  />',
+        out,
+        count=1,
+    )
+    out = _OG_URL_RE.sub(
+        f'<meta property="og:url" content="{_esc(canonical)}" />',
+        out,
+        count=1,
+    )
+    out = _TW_TITLE_RE.sub(
+        f'<meta name="twitter:title" content="{_esc(title)}" />',
+        out,
+        count=1,
+    )
+    out = _TW_DESC_RE.sub(
+        f'<meta\n    name="twitter:description"\n    content="{_esc(description)}"\n  />',
+        out,
+        count=1,
+    )
+    out = _JSONLD_RE.sub(
+        _jsonld_website(origin, title, description, canonical),
+        out,
+        count=1,
+    )
+    aside = (
+        f'<aside id="seo-prerender" class="seo-prerender">\n'
+        f"    {seo_body_html}\n"
+        f"  </aside>"
+    )
+    out = _SEO_ASIDE_RE.sub(aside, out, count=1)
+    return out
+
+
+def write_text(path: Path, text: str) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(text, encoding="utf-8", newline="\n")
+
+
+def build_sitemap(urls: list[str], origin: str) -> str:
+    lines = [
+        '<?xml version="1.0" encoding="UTF-8"?>',
+        '<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">',
+    ]
+    for path in urls:
+        loc = xml_escape(_abs(origin, path))
+        lines.append("  <url>")
+        lines.append(f"    <loc>{loc}</loc>")
+        lines.append("  </url>")
+    lines.append("</urlset>")
+    lines.append("")
+    return "\n".join(lines)
+
+
+def build_llms_txt(urls: list[tuple[str, str]], origin: str) -> str:
+    lines = [
+        "# 上分帝 (Climperor)",
+        "",
+        "> Dota 2 英雄机制克制与搭档、段位走势、机制查询、物品与版本更新。",
+        "",
+        f"站点：{_abs(origin, '/')}",
+        "",
+        "## 主要页面",
+        "",
+    ]
+    for path, title in urls:
+        lines.append(f"- [{title}]({_abs(origin, path)})")
+    lines.append("")
+    lines.append("## 说明")
+    lines.append("")
+    lines.append("- 克制/搭档为定性机制边（含理由），不是胜率因果结论。")
+    lines.append("- 走势/对位数据来自公开统计源，页面会标注窗口与段位。")
+    lines.append("- 完整交互界面面向桌面浏览器。")
+    lines.append("")
+    return "\n".join(lines)
+
+
+def write_seo_bundle(
+    out: Path,
+    template_html: str,
+    payload: dict,
+    *,
+    site_origin: str = DEFAULT_SITE_ORIGIN,
+) -> dict[str, int]:
+    """Write prerendered pages + sitemap.xml + llms.txt into ``out``.
+
+    Root ``index.html`` is rewritten in place with home SEO. Nested pages
+    are written as ``heroes/<key>/index.html`` and ``mechanics/<effect>/index.html``.
+    """
+    origin = (site_origin or DEFAULT_SITE_ORIGIN).rstrip("/")
+    sitemap_paths: list[str] = []
+    llms_entries: list[tuple[str, str]] = []
+    counts = {"top": 0, "heroes": 0, "mechanics": 0}
+
+    for path, title, desc in TOP_PAGES:
+        full_title = f"{title} — 上分帝"
+        body = _top_seo_body(path, title, desc, payload)
+        html_doc = inject_seo(
+            template_html,
+            title=full_title,
+            description=desc,
+            canonical=_abs(origin, path if path != "/" else "/"),
+            seo_body_html=body,
+            origin=origin,
+        )
+        if path == "/":
+            write_text(out / "index.html", html_doc)
+        else:
+            # /heroes → heroes/index.html etc.
+            rel = path.strip("/")
+            write_text(out / rel / "index.html", html_doc)
+        sitemap_paths.append(path if path != "/" else "/")
+        llms_entries.append((path if path != "/" else "/", title))
+        counts["top"] += 1
+
+    for hero in payload.get("heroes") or []:
+        if not isinstance(hero, dict):
+            continue
+        key = hero.get("key")
+        if not key:
+            continue
+        key = str(key)
+        name = str(hero.get("name_loc") or key)
+        aliases = [str(a) for a in (hero.get("aliases") or []) if a]
+        alias_bit = f"（{'、'.join(aliases[:3])}）" if aliases else ""
+        title = f"{name} 克制与搭档 — 上分帝"
+        desc = (
+            f"{name}{alias_bit}的 Dota 2 机制克制、被克制与搭档参考，"
+            f"以及技能、出装与走势（上分帝）。"
+        )
+        path = f"/heroes/{key}"
+        html_doc = inject_seo(
+            template_html,
+            title=title,
+            description=desc,
+            canonical=_abs(origin, path),
+            seo_body_html=_hero_seo_body(hero, payload),
+            origin=origin,
+        )
+        write_text(out / "heroes" / key / "index.html", html_doc)
+        sitemap_paths.append(path)
+        llms_entries.append((path, f"{name} 克制与搭档"))
+        counts["heroes"] += 1
+
+    mq = payload.get("mechanic_query") or {}
+    labels = mq.get("labels") or {}
+    for effect in mq.get("order") or []:
+        effect = str(effect)
+        label = labels.get(effect) or effect
+        # Default effect is bare /mechanics (already written as top page).
+        if effect == "basic_dispel":
+            continue
+        path = f"/mechanics/{effect}"
+        blurb = (mq.get("blurbs") or {}).get(effect) or f"查询施加「{label}」的技能与物品。"
+        title = f"{label} — 机制查询 — 上分帝"
+        html_doc = inject_seo(
+            template_html,
+            title=title,
+            description=str(blurb),
+            canonical=_abs(origin, path),
+            seo_body_html=_mechanic_seo_body(effect, payload),
+            origin=origin,
+        )
+        write_text(out / "mechanics" / effect / "index.html", html_doc)
+        sitemap_paths.append(path)
+        llms_entries.append((path, f"{label}（机制）"))
+        counts["mechanics"] += 1
+
+    # Prefer stable order: tops first, then heroes, then mechanics (already).
+    write_text(out / "sitemap.xml", build_sitemap(sitemap_paths, origin))
+    # Keep llms.txt focused: tops + sample of heroes would be huge; include all
+    # tops + mechanics + first-line note that hero URLs follow /heroes/{key}.
+    llms_compact = [(p, t) for p, t in llms_entries if not p.startswith("/heroes/")]
+    llms_compact.append(("/heroes/{key}", "各英雄克制/搭档页（key 为英雄英文键）"))
+    write_text(out / "llms.txt", build_llms_txt(llms_compact, origin))
+    return counts
@@ -1115,11 +1115,29 @@ class Handler(BaseHTTPRequestHandler):
            return
        rel = path.lstrip("/")
        candidate = (WEB_DIR / rel).resolve()
-        if not str(candidate).startswith(str(WEB_DIR.resolve())) or not candidate.is_file():
-            self.send_error(404)
+        web_root = WEB_DIR.resolve()
+        if str(candidate).startswith(str(web_root)) and candidate.is_file():
+            ctype = mimetypes.guess_type(str(candidate))[0] or "application/octet-stream"
+            self._send(200, candidate.read_bytes(), ctype)
            return
-        ctype = mimetypes.guess_type(str(candidate))[0] or "application/octet-stream"
-        self._send(200, candidate.read_bytes(), ctype)
+        # History SPA fallback: /heroes/axe → index.html (client router).
+        spa_pages = {
+            "heroes",
+            "rankings",
+            "matches",
+            "streamers",
+            "trends",
+            "mechanics",
+            "items",
+            "patches",
+        }
+        first = path.strip("/").split("/", 1)[0] if path.strip("/") else ""
+        if first in spa_pages:
+            index = WEB_DIR / "index.html"
+            if index.is_file():
+                self._send(200, index.read_bytes(), "text/html; charset=utf-8")
+                return
+        self.send_error(404)

    def do_POST(self) -> None:  # noqa: N802
        path = urlparse(self.path).path