feat(web): simplify product language

docs/en/benchmarks/index.md

@@ -36,7 +36,7 @@ benchmark:
         accessible: 86.73 percent task resolution
         metric: tasks resolved with PowerContext on
   orientation:
-    title: Two benchmarks. Two different questions.
+    title: Two benchmarks, two questions.
     lead: "Memory quality matters twice: first when an agent must recover what happened, then when it must use context to finish real work."
     tests:
       - name: LoCoMo
@@ -128,7 +128,7 @@ benchmark:
     scope_title: What this result covers
     scope: The 90.78% result is 1,398 correct answers from 1,540 questions in categories 1-4. It does not claim results for LoCoMo event summarization or multimodal dialogue generation.
   swe:
-    title: SWE-bench Pro tests whether context changes the patch.
+    title: Does context improve the patch?
     lead: Each task starts from a real codebase and issue. Codex edits the repository, and the official task tests decide whether the patch resolves the problem.
     method:
       - title: Same task set
@@ -422,8 +422,10 @@ benchmark:
           score: 1.51%
           ci: ±1.51
   reading:
-    title: Read each result for the question it answers.
+    title: Compare what each result measures.
     lead: The two evaluations share a context theme, but their inputs, outputs, and graders are intentionally different.
+    scroll_hint: Swipe sideways to compare both evaluations.
+    table_label: Comparison of the LoCoMo and SWE-bench Pro evaluations
     columns:
       dimension: Evaluation dimension
     rows:

docs/en/docs/index.md

@@ -1,20 +1,22 @@
 ---
 template: docs-overview.html
-title: Choose an Agent or API and complete the first loop
+title: Choose an Agent or API
 description: Add the HTTP Memory API to an existing AI application, or configure an Agent and verify Memory and Handoff.
 page_type: docs-overview
 overview:
   intro: If you are new to PowerContext, choose the entry that matches your application. Use HTTP directly from an existing AI application, or start with the Agent quickstart for Codex, Claude Code, OpenCode, and other hosts.
   sections:
-    - title: Recommended learning path
-      description: Complete a local API or Agent loop first, then connect another host or adopt the complete work workflow.
+    - title: Start here
+      description: Complete an API or Agent quickstart first. The remaining guides help you continue in the Agent you use.
       cards:
         - title: HTTP API lifecycle tutorial
           description: Use one small Python flow to connect context, Experience, Skill, and Review; use Scalar for endpoint reference.
           href: en/docs/tutorials/api-quickstart/
+          featured: true
         - title: Agent step-by-step quickstart
           description: Choose a host, install and diagnose it, complete Memory, then use its one-line, pc_*, or /pc Handoff path.
           href: en/docs/tutorials/agent-quickstart/
+          featured: true
         - title: Complete Codex tutorial
           description: Go deeper into the Codex Hook, MCP Skill, Memory lifecycle, and one-line durable Handoff.
           href: en/docs/tutorials/codex-quickstart/

docs/en/index.md

@@ -1,50 +1,44 @@
 ---
 template: home.html
 title: PowerContext
-description: Add Memory to your own AI over HTTP, or choose a supported agent for cross-session recovery and Handoff.
+description: Continue work in a new session without restating decisions, constraints, and progress. Use a supported Agent or connect your own AI over HTTP.
 hide:
   - navigation
   - toc
   - footer
 home:
   hero:
-    label: Open source · Project scoped · Local by default
+    label: Open source. Local storage by default.
     title:
-      - Start a new session.
-      - Keep moving.
-    lead: PowerContext keeps decisions, constraints, next steps, and work boundaries outside the chat. Existing AI applications can call HTTP directly; Codex, Claude Code, DSH, OpenCode, and other agents can connect to the same Server.
-    note: You control the model, identity, and write authorization; PowerContext provides cited Memory, request-time context, and history-preserving revisions.
+      - Pick up where you left off.
+    lead: Save the decisions, constraints, and progress you will need later in PowerContext. Find them again in another session.
     actions:
-      - label: Choose your agent
+      - label: Start with an Agent
         href: en/docs/tutorials/agent-quickstart/
         kind: primary
-      - label: Add the API to your AI
+      - label: Connect your AI over HTTP
         href: en/docs/tutorials/api-quickstart/
         kind: secondary
   continuity:
-    label: One project, multiple sessions
-    title: One agent stops. Work continues.
-    lead: Different agents can implement, review, and validate compatibility. Handoff transfers an inspected boundary; a human decides whether to continue.
+    title: No need to repeat yourself.
+    lead: Save information that needs to last. Later sessions can use it and see how it changed.
     steps:
-      - title: Agent A implements
-        description: Record decisions, constraints, and next steps, then assemble the objective, changes, checks, and omissions as a Handoff.
-      - title: Agent B checks independently
-        description: Receive the exact Revision in another host and verify its evidence and risks against the current repository.
-      - title: A human decides
-        description: Confirm scope, capability, and authorization, then continue, request clarification, or decline. Receipt is not completion.
+      - title: Save
+        description: Record the decisions, constraints, and next steps you will need later.
+      - title: Find
+        description: Find the relevant information in another Agent or application without pasting content you already saved.
+      - title: Continue
+        description: Check that the information still fits the current situation, then continue the work.
   ownership:
-    label: Memory and Handoff
     title:
-      - Keep what lasts.
+      - Save what matters.
       - Hand off the work.
-    lead: Memory keeps decisions, constraints, conventions, and next steps in a searchable history. Revise or retire an entry without losing the record.
-    handoff: A Handoff captures the current objective, verified progress, blockers, and next action. Commit it when the work becomes a project milestone.
-    result: "LOCOMO: 90.78% correct · 1.38 s p95 search latency"
-    command: powercontext server run
+    lead: Memory keeps information that stands on its own and will matter later. You can revise or retire an entry while preserving its history.
+    handoff: A Handoff saves the current objective and progress. It also records blockers and the next action so another person, Agent, or later session can continue. A proven approach can become an Experience. Recurring work can become a Skill.
     primary_action:
-      label: Open the Agent quickstart
+      label: Read the Agent quickstart
       href: en/docs/tutorials/agent-quickstart/
     secondary_action:
-      label: Open the API lifecycle tutorial
+      label: Connect your AI over HTTP
       href: en/docs/tutorials/api-quickstart/
 ---

docs/zh/benchmarks/index.md

@@ -10,8 +10,7 @@ benchmark:
   hero:
     label: 评估测试证据
     title:
-      - 把上下文能力，
-      - 交给评估测试。
+      - 上下文压力测试
     lead: 两项公开评测，分别检验长程记忆与真实仓库软件工程，并给出可核对的结果。
     actions_label: 跳转到具体评测
     actions:
@@ -36,7 +35,7 @@ benchmark:
         accessible: 任务解决率 86.73%
         metric: 开启 PowerContext 后的任务解决率
   orientation:
-    title: 两项评估测试，回答两个不同问题。
+    title: 两项评测，两个问题。
     lead: Agent 先要找回发生过什么，再要把上下文用于真实工作。两项评测分别检验这两层能力。
     tests:
       - name: LoCoMo
@@ -128,7 +127,7 @@ benchmark:
     scope_title: 结果覆盖范围
     scope: 90.78% 来自类别 1-4 的 1,540 个问题，其中答对 1,398 个。该结果不代表 LoCoMo 的事件总结或多模态对话生成任务。
   swe:
-    title: SWE-bench Pro 检验上下文是否改变补丁结果。
+    title: 上下文能改善补丁结果吗？
     lead: 每个任务都从真实代码库与 Issue 开始。Codex 修改仓库，再由任务自带的正式测试判断补丁是否解决问题。
     method:
       - title: 相同任务集
@@ -422,8 +421,10 @@ benchmark:
           score: 1.51%
           ci: ±1.51
   reading:
-    title: 用每项结果回答它真正评测的问题。
+    title: 先看清每项结果测什么。
     lead: 两项评测都与上下文有关，但输入、输出与评分方式不同，不能把两个分数直接横向比较。
+    scroll_hint: 左右滑动，比较两项评测。
+    table_label: LoCoMo 与 SWE-bench Pro 评测对比
     columns:
       dimension: 评测维度
     rows:

docs/zh/docs/index.md

@@ -1,20 +1,22 @@
 ---
 template: docs-overview.html
-title: 选择 Agent 或 API，跑通第一个闭环
+title: 选择 Agent 或 API
 description: 为现有 AI 应用接入 HTTP Memory API，或为选定 Agent 配置集成并验证 Memory 与 Handoff。
 page_type: docs-overview
 overview:
   intro: 第一次使用时，先选择自己的入口。已有 AI 应用可以直接使用 HTTP API；使用 Codex、Claude Code、OpenCode 等 Host 时，从 Agent 分步入门开始。
   sections:
-    - title: 推荐学习路径
-      description: 先通过 API 或 Agent 跑通本地闭环，再按需要接入其他 Host 或更完整的工作流程。
+    - title: 从这里开始
+      description: 先通过 API 或 Agent 跑通一次。其余说明帮助你在所用 Agent 中继续工作。
       cards:
         - title: HTTP API 生命周期教程
           description: 用一个小型 Python 流程接入上下文、Experience、Skill 与 Review；接口参考统一使用 Scalar。
           href: zh/docs/tutorials/api-quickstart/
+          featured: true
         - title: Agent 分步入门
           description: 选择 Host，安装并诊断集成，跑通 Memory，再按一句话、pc_* 或 /pc 路径完成 Handoff。
           href: zh/docs/tutorials/agent-quickstart/
+          featured: true
         - title: Codex 完整教程
           description: 深入 Codex Hook、MCP Skill、Memory 生命周期与一句话 durable Handoff。
           href: zh/docs/tutorials/codex-quickstart/

docs/zh/index.md

@@ -1,50 +1,45 @@
 ---
 template: home.html
 title: PowerContext
-description: 通过 HTTP 为自己的 AI 接入 Memory，或选择受支持的 Agent 跑通跨会话恢复与 Handoff。
+description: 换个会话继续工作，不必重新交代决定、约束和进展。可以使用受支持的 Agent，也可以通过 HTTP 接入自己的 AI。
 hide:
   - navigation
   - toc
   - footer
 home:
   hero:
-    label: 开源 · 项目级 · 默认本地
+    label: 开源，默认使用本地存储
     title:
-      - 换一个会话，
-      - 项目继续向前。
-    lead: PowerContext 把决定、约束、下一步和任务边界保存在对话之外。现有 AI 应用可以直接调用 HTTP API；Codex、Claude Code、DSH、OpenCode 等 Agent 也可以连接同一个 Server。
-    note: 你控制模型、身份和写入授权；PowerContext 提供带 citation 的 Memory、请求时上下文和保留历史的修订。
+      - 换个会话，
+      - 接着做。
+    lead: 把需要保留的决定、约束和进展存进 PowerContext。换个会话，也能找回这些信息。
     actions:
-      - label: 选择你的 Agent
+      - label: 从 Agent 开始
         href: zh/docs/tutorials/agent-quickstart/
         kind: primary
-      - label: 为自己的 AI 接入 API
+      - label: 通过 HTTP 接入 AI
         href: zh/docs/tutorials/api-quickstart/
         kind: secondary
   continuity:
-    label: 同一项目，多个会话
-    title: 一个 Agent 停下，工作仍能继续。
-    lead: 实现、独立检查和兼容性验证可以由不同 Agent 完成；Handoff 传递经过检查的边界，人决定是否继续。
+    title: 重要信息不必反复交代。
+    lead: 把需要长期保留的信息记下来。后续会话可以继续使用，也能查看修改记录。
     steps:
-      - title: Agent A 实现
-        description: 记录项目决定、约束和下一步，并把目标、改动、检查与遗漏整理为 Handoff。
-      - title: Agent B 独立核对
-        description: 在另一个 Host 中按 exact Revision 接收，再用当前仓库重新验证证据和风险。
-      - title: 人决定
-        description: 人确认范围、能力和授权，决定继续、澄清或拒绝；接收不等于任务已经完成。
+      - title: 记下
+        description: 记下以后还要用的决定、约束和下一步。
+      - title: 找回
+        description: 换到另一个 Agent 或应用时，找到相关信息，不必重新粘贴已经保存的内容。
+      - title: 继续
+        description: 确认这些信息仍符合当前情况，然后接着做。
   ownership:
-    label: Memory 与 Handoff
     title:
-      - 留下长期信息，
-      - 交接当前工作。
-    lead: Memory 保存决定、约束、约定和下一步，并保留可检索的历史。修订或停用条目，不会丢失记录。
-    handoff: Handoff 记录当前目标、已验证进展、阻塞项和下一步行动。工作形成项目里程碑后再提交。
-    result: "LOCOMO：答对率 90.78% · 搜索 p95 延迟 1.38 秒"
-    command: powercontext server run
+      - 记住以后还要用的，
+      - 交接手头正在做的。
+    lead: 记忆保存以后仍会用到、单独拿出来也能理解的信息。更新或停用其中一条，不会丢失历史版本。
+    handoff: 交接保存当前目标和进度，也说明哪里卡住、下一步做什么。换人、换 Agent 或稍后再做时，可以从这里继续。已经验证的做法可以整理成经验，需要反复执行的工作可以整理成技能。
     primary_action:
-      label: 打开 Agent 分步入门
+      label: 阅读 Agent 快速入门
       href: zh/docs/tutorials/agent-quickstart/
     secondary_action:
-      label: 打开 HTTP API 生命周期教程
+      label: 通过 HTTP 接入 AI
       href: zh/docs/tutorials/api-quickstart/
 ---