fix(ui): tarifs publics unifiés et prix annuel lisible

La page tarifs reprend la barre du site public. En annuel, Pro affiche
8,25 € par mois (99 € l’année), plus le 99 € comme gros chiffre.

Co-authored-by: Cursor <cursoragent@cursor.com>

memento-note/locales/zh.json

@@ -2190,7 +2190,7 @@
       "custom": "自定义"
     },
     "typeDescriptions": {
-      "scraper": "抓取多个网站并创建摘要",
+      "scraper": "阅读多个网站并写出摘要",
       "researcher": "搜索有关主题的信息",
       "monitor": "监视笔记本并分析笔记",
       "slideGenerator": "根据笔记创建 PowerPoint 演示文稿",
@@ -2203,7 +2203,7 @@
       "namePlaceholder": "例如：周二 AI 观察",
       "description": "描述（可选）",
       "descriptionPlaceholder": "每周 AI 新闻摘要",
-      "urlsLabel": "要抓取的 URL",
+      "urlsLabel": "要阅读的页面地址",
       "urlsOptional": "（可选）",
       "sourceNotebook": "要监视的笔记本",
       "selectNotebook": "选择笔记本...",
@@ -2248,7 +2248,7 @@
       "notifyEmail": "邮件通知",
       "notifyEmailHint": "每次运行后通过邮件接收代理结果",
       "includeImages": "包含图片",
-      "includeImagesHint": "从抓取的页面中提取图片并附加到生成的笔记",
+      "includeImagesHint": "从已阅读的页面取出图片并附到笔记",
       "back": "返回",
       "configuration": "配置",
       "options": "选项",
@@ -2347,15 +2347,15 @@
       },
       "veilleAI": {
         "name": "AI 观察",
-        "description": "抓取 5 个 AI 专业网站并生成每周摘要。"
+        "description": "阅读 5 个 AI 网站并写出每周摘要。"
       },
       "veilleTech": {
         "name": "科技观察",
-        "description": "抓取主要科技网站并创建新闻摘要。"
+        "description": "阅读主要科技网站并写出新闻摘要。"
       },
       "veilleDev": {
         "name": "开发观察",
-        "description": "抓取开发网站并总结新技术和框架。"
+        "description": "阅读开发网站并总结新技术。"
       },
       "surveillant": {
         "name": "笔记观察者",
@@ -2402,7 +2402,7 @@
     "tools": {
       "title": "代理工具",
       "webSearch": "网络搜索",
-      "webScrape": "网页抓取",
+      "webScrape": "阅读网页",
       "noteSearch": "笔记搜索",
       "noteRead": "读取笔记",
       "noteCreate": "创建笔记",
@@ -2431,15 +2431,15 @@
       "btnLabel": "帮助",
       "close": "关闭",
       "whatIsAgent": "什么是代理？",
-      "whatIsAgentContent": "An **agent** is an AI assistant that runs automatically to perform tasks for you. It has access to **tools** (web search, web scraping, note reading...) and produces a **note** with its results.\n\nThink of it as a small autonomous worker: you give it a mission, it researches or scrapes information, then writes a structured note you can read later.",
+      "whatIsAgentContent": "An **agent** is an AI assistant that runs automatically to perform tasks for you. It has access to **tools** (web search, reading pages, note reading...) and produces a **note** with its results.\n\nThink of it as a small autonomous worker: you give it a mission, it researches or reads pages, then writes a structured note you can read later.",
       "howToUse": "如何使用代理？",
       "howToUseContent": "1. 点击**\"新建智能体\"**（或从页面底部的**模板**开始）。",
       "types": "代理类型",
-      "typesContent": "### Researcher\nSearches the web on a **topic you define** and creates a structured note with sources and references.\n\n- **Fields:** name, research topic (e.g. \"Latest advances in quantum computing\")\n- **Default tools:** web search, web scraping, note search, note creation\n- **Requirements:** a web search provider must be configured (SearXNG or Brave Search)\n\n### Monitor (Scraper)\nScrapes a **list of URLs** you specify and produces a summary of their content.\n\n- **Fields:** name, list of URLs (e.g. tech news sites, blogs...)\n- **Default tools:** web scraping, note creation\n- **Use case:** weekly tech watch, competitor monitoring, blog roundups\n\n### Observer (Notebook Monitor)\nReads notes from a **notebook you select** and produces analysis, connections, and suggestions.\n\n- **Fields:** name, source notebook (the one to analyze)\n- **Default tools:** note search, note read, note creation\n- **Use case:** find connections between your notes, get reading suggestions, detect recurring themes\n\n### Custom\nA blank canvas: you write your own **prompt** and pick your own **tools**.\n\n- **Fields:** name, description, custom instructions (in Advanced mode)\n- **No default tools** — you choose exactly what the agent needs\n- **Use case:** anything creative or specific that doesn't fit the other types",
+      "typesContent": "### Researcher\nSearches the web on a **topic you define** and creates a structured note with sources and references.\n\n- **Fields:** name, research topic (e.g. \"Latest advances in quantum computing\")\n- **Default tools:** web search, reading pages, note search, note creation\n- **Requirements:** a web search provider must be configured (SearXNG or Brave Search)\n\n### Monitor\nReads a **list of pages** you give it and writes a summary.\n\n- **Fields:** name, list of URLs (e.g. tech news sites, blogs...)\n- **Default tools:** reading pages, note creation\n- **Use case:** weekly tech watch, competitor monitoring, blog roundups\n\n### Observer (Notebook Monitor)\nReads notes from a **notebook you select** and produces analysis, connections, and suggestions.\n\n- **Fields:** name, source notebook (the one to analyze)\n- **Default tools:** note search, note read, note creation\n- **Use case:** find connections between your notes, get reading suggestions, detect recurring themes\n\n### Custom\nA blank canvas: you write your own **prompt** and pick your own **tools**.\n\n- **Fields:** name, description, custom instructions (in Advanced mode)\n- **No default tools** — you choose exactly what the agent needs\n- **Use case:** anything creative or specific that doesn't fit the other types",
       "advanced": "高级模式（AI指令，最大迭代）",
       "advancedContent": "点击表单底部的**\"高级模式\"**以访问附加设置。",
       "tools": "可用工具（详细）",
-      "toolsContent": "When advanced mode is enabled, you can choose exactly which tools the agent can use.\n\n### Web Search\nAllows the agent to **search the internet** via SearXNG or Brave Search.\n\n- **What it does:** The agent formulates a query, gets search results, and can then scrape the most relevant pages.\n- **When to enable:** When the agent needs to find information on a topic (Researcher or Custom type).\n- **Configuration required:** SearXNG (with JSON format enabled) or a Brave Search API key. Configurable in **Admin > Agent Tools**.\n- **Example:** The agent searches \"React Server Components best practices 2025\", gets 10 results, then scrapes the top 3.\n\n### Web Scrape\nAllows the agent to **extract text content from a web page** given its URL.\n\n- **What it does:** The agent visits a URL and retrieves the structured text (headings, paragraphs, lists). Ads, menus and footers are typically filtered out.\n- **When to enable:** For the Monitor type (mandatory), or any agent that needs to read web pages.\n- **Configuration:** Works out of the box, but a **Jina Reader API key** improves quality and removes rate limits. Configurable in **Admin > Agent Tools**.\n- **Example:** The agent scrapes 5 tech blogs and produces a synthesized summary.\n\n### Note Search\nAllows the agent to **search your existing notes**.\n\n- **What it does:** The agent performs a text search across all your notes (or a specific notebook).\n- **When to enable:** For Observer-type agents, or any agent that needs to cross-reference information with your notes.\n- **Configuration:** None — works immediately.\n- **Example:** The agent searches all notes containing \"machine learning\" to see what you've already written on the topic.\n\n### Read Note\nAllows the agent to **read the full content of a specific note**.\n\n- **What it does:** After finding a note (via Note Search), the agent can read its entire content to analyze or use it.\n- **When to enable:** As a companion to Note Search. Enable both together so the agent can search AND read.\n- **Configuration:** None.\n- **Example:** The agent finds 5 notes about \"productivity\", reads them all, and writes a synthesis.\n\n### Create Note\nAllows the agent to **write a new note** in your target notebook.\n\n- **What it does:** The agent creates a note with a title and content. This is how results end up in your notebooks.\n- **When to enable:** Almost always — without this tool, the agent cannot save its results. **Leave it enabled by default.**\n- **Configuration:** None.\n- **Example:** The agent creates a note \"Tech Watch - Week 16\" with a summary of 5 articles.\n\n### Fetch URL\nAllows the agent to **download the raw content of a URL** (HTML, JSON, text...).\n\n- **What it does:** Unlike scraping which extracts clean text, Fetch URL retrieves raw content. Useful for APIs, JSON files, or non-standard pages.\n- **When to enable:** When the agent needs to query REST APIs, read RSS feeds, or access raw data.\n- **Configuration:** None.\n- **Example:** The agent queries the GitHub API to list the latest commits of a project.\n\n### Memory\nAllows the agent to **access its previous execution history**.\n\n- **What it does:** The agent can search through results from past runs. This lets it compare, track changes, or avoid repeating the same information.\n- **When to enable:** For agents that run regularly and need to maintain continuity between executions.\n- **Configuration:** None.\n- **Example:** The agent compares this week's news with last week's and highlights what's new.",
+      "toolsContent": "When advanced mode is enabled, you can choose exactly which tools the agent can use.\n\n### Web Search\nAllows the agent to **search the internet** via SearXNG or Brave Search.\n\n- **What it does:** The agent formulates a query, gets search results, then can read the most useful pages.\n- **When to enable:** When the agent needs to find information on a topic (Researcher or Custom type).\n- **Configuration required:** SearXNG (with JSON format enabled) or a Brave Search API key. Configurable in **Admin > Agent Tools**.\n- **Example:** The agent searches \"React Server Components best practices 2025\", gets 10 results, then reads the top 3.\n\n### Read web pages\nAllows the agent to **read the text of a page** from its address.\n\n- **What it does:** The agent visits a URL and retrieves the structured text (headings, paragraphs, lists). Ads, menus and footers are typically filtered out.\n- **When to enable:** For the Monitor type (mandatory), or any agent that needs to read web pages.\n- **Configuration:** Works out of the box, but a **Jina Reader API key** improves quality and removes rate limits. Configurable in **Admin > Agent Tools**.\n- **Example:** The agent scrapes 5 tech blogs and produces a synthesized summary.\n\n### Note Search\nAllows the agent to **search your existing notes**.\n\n- **What it does:** The agent performs a text search across all your notes (or a specific notebook).\n- **When to enable:** For Observer-type agents, or any agent that needs to cross-reference information with your notes.\n- **Configuration:** None — works immediately.\n- **Example:** The agent searches all notes containing \"machine learning\" to see what you've already written on the topic.\n\n### Read Note\nAllows the agent to **read the full content of a specific note**.\n\n- **What it does:** After finding a note (via Note Search), the agent can read its entire content to analyze or use it.\n- **When to enable:** As a companion to Note Search. Enable both together so the agent can search AND read.\n- **Configuration:** None.\n- **Example:** The agent finds 5 notes about \"productivity\", reads them all, and writes a synthesis.\n\n### Create Note\nAllows the agent to **write a new note** in your target notebook.\n\n- **What it does:** The agent creates a note with a title and content. This is how results end up in your notebooks.\n- **When to enable:** Almost always — without this tool, the agent cannot save its results. **Leave it enabled by default.**\n- **Configuration:** None.\n- **Example:** The agent creates a note \"Tech Watch - Week 16\" with a summary of 5 articles.\n\n### Fetch URL\nAllows the agent to **download the raw content of a URL** (HTML, JSON, text...).\n\n- **What it does:** Unlike scraping which extracts clean text, Fetch URL retrieves raw content. Useful for APIs, JSON files, or non-standard pages.\n- **When to enable:** When the agent needs to query REST APIs, read RSS feeds, or access raw data.\n- **Configuration:** None.\n- **Example:** The agent queries the GitHub API to list the latest commits of a project.\n\n### Memory\nAllows the agent to **access its previous execution history**.\n\n- **What it does:** The agent can search through results from past runs. This lets it compare, track changes, or avoid repeating the same information.\n- **When to enable:** For agents that run regularly and need to maintain continuity between executions.\n- **Configuration:** None.\n- **Example:** The agent compares this week's news with last week's and highlights what's new.",
       "frequency": "频率和计划",
       "frequencyContent": "| 频率 | 行为\n|-----------|----------\n| **手动** | 您自己点击\"运行\"。",
       "targetNotebook": "目标笔记本",
@@ -2447,7 +2447,7 @@
       "templates": "模板",
       "templatesContent": "Templates are pre-configured agents ready to install in one click. You'll find them at the **bottom of the Agents page**.\n\nAvailable templates include:\n\n- **AI Watch** — weekly AI news roundup from 5 specialized sites\n- **Tech Watch** — general tech news summary\n- **Dev Watch** — developer news and new frameworks\n- **Note Observer** — analyzes a notebook and suggests connections\n- **Topic Researcher** — deep research on a specific topic\n\nOnce installed, you can edit the agent to customize it.",
       "tips": "提示和故障排除",
-      "tipsContent": "- **Start with a template** and customize it — it's the fastest way to get a working agent\n- **Test with \"Manual\"** frequency before enabling automatic scheduling\n- **A \"Researcher\" agent requires a web search provider** — configure SearXNG (JSON format) or Brave Search in **Admin > Agent Tools**\n- **If an agent fails**, click on its card then **History** to see the execution log and tool traces\n- **The \"Enabled/Disabled\" toggle** lets you pause an agent without deleting it\n- **Web scraping quality** improves with a Jina Reader API key (optional, in Admin > Agent Tools)\n- **Combine \"Note Search\" + \"Read Note\"** so the agent can find AND analyze your notes' content\n- **Enable \"Memory\"** if your agent runs regularly — it will avoid repeating the same information across runs",
+      "tipsContent": "- **Start with a template** and customize it — it's the fastest way to get a working agent\n- **Test with \"Manual\"** frequency before enabling automatic scheduling\n- **A \"Researcher\" agent requires a web search provider** — configure SearXNG (JSON format) or Brave Search in **Admin > Agent Tools**\n- **If an agent fails**, click on its card then **History** to see the execution log and tool traces\n- **The \"Enabled/Disabled\" toggle** lets you pause an agent without deleting it\n- **Page-reading quality** improves with a Jina Reader API key (optional, in Admin > Agent Tools)\n- **Combine \"Note Search\" + \"Read Note\"** so the agent can find AND analyze your notes' content\n- **Enable \"Memory\"** if your agent runs regularly — it will avoid repeating the same information across runs",
       "tooltips": {
         "agentType": "选择代理将执行的任务类型。每种类型具有不同的功能和字段。",
         "researchTopic": "代理将在网络上研究的主题。请具体说明以获得更好的结果。",
@@ -3176,7 +3176,8 @@
     "fetchStatusFailed": "无法获取计费状态",
     "fetchQuotasFailed": "无法获取配额",
     "fetchInvoicesFailed": "无法加载计费历史记录。",
-    "savePercent": "节省 ~17%",
+    "savePercent": "节省约 {percent}%",
+    "billedYearTotal": "即每年 {price}",
     "cancelSubscription": "取消订阅",
     "changeOffer": "更换套餐",
     "downgradeToFree": "回到免费套餐",
@@ -3385,7 +3386,7 @@
         "feature5": "安装陪同"
       },
       "basicPrice": "免费",
-      "savePercent": "节省约 17%",
+      "savePercent": "节省约 {percent}%",
       "proMonthly": "€9.90",
       "proAnnualMonthly": "€8.25",
       "businessMonthly": "€29.90",
@@ -3665,7 +3666,7 @@
     "mappingHint": "这可能需要一到三分钟。您可以继续浏览；页面会自动更新。",
     "analyzeNow": "更新主题",
     "emptyNeedMoreNotes": "再添加 {count} 条笔记即可分组主题（至少 10 条）。",
-    "embeddingsHint": "仅{indexed}/{total}笔记被AI索引。",
+    "embeddingsHint": "仅 {indexed} / {total} 条笔记已准备好按主题分组。",
     "vsGraphHint": "这与\"链接地图\"不同：这里AI按语义分组，而非按链接。",
     "openGraphMap": "打开链接地图",
     "analysisFailed": "分析失败。请检查AI设置。",
@@ -4492,7 +4493,7 @@
     "convertSuccess": "转换完成！已创建关联笔记本。",
     "convertToNotebook": "转换为笔记本",
     "converting": "转换中…",
-    "createLocalDb": "创建独立本地数据库",
+    "createLocalDb": "在本笔记中创建表格",
     "createNotebook": "创建笔记本",
     "defaultOption1": "选项 1",
     "defaultOption2": "选项 2",
@@ -4514,7 +4515,7 @@
     "keywordMatch": "关键词",
     "linkToNotebook": "链接到笔记本",
     "loadError": "加载结构化数据失败。",
-    "localDbTitle": "独立数据库",
+    "localDbTitle": "本笔记中的表格",
     "namePlaceholder": "输入名称…",
     "noEchoFound": "未找到相近笔记。",
     "noNotebook": "此块需要笔记本。先将此笔记移至笔记本。",
@@ -4528,8 +4529,8 @@
     "selectNotebook": "链接到笔记本",
     "selectOptionsPlaceholder": "用逗号分隔的选项",
     "semanticEcho": "语义共振",
-    "switchToLocalDb": "切换到本地数据库",
-    "turnIntoLabel": "内联数据库",
+    "switchToLocalDb": "回到本笔记的表格",
+    "turnIntoLabel": "笔记中的表格",
     "untitled": "无标题"
   },
   "structuredViews": {