Visibility release candidate: human authority and public evolution

AI_ENTRYPOINT.md

@@ -40,6 +40,8 @@ Validate LabNote by its expected structure and entrypoint, not by a fixed reposi
 
 No current-run visitor handle, no write.
 
+An AI may identify material that could be useful to retain and may explain or draft a proposed deposit. It must not start, create, change, or register a project record merely because it considers that material useful. A human operator must directly initiate or approve the specific write in the current interaction.
+
 Do not reuse visitor handles, branches, storage locations, or permissions from earlier conversation context unless the human operator explicitly confirms them for this run.
 
 If source reports/documents are pasted or uploaded in the current prompt, treat that material as the source material.

README.md

@@ -6,7 +6,7 @@ CapstanAI LabNote is a lightweight, file-based project ledger for people working
 
 You decide what belongs in the record and what requires review. A shared entry route gives each AI a clear reading path, defined places to leave work and stopping points for human decisions. Basic ledger use needs no LabNote background service, database or model API key.
 
-CapstanAI is the project name; LabNote is its practical ledger.
+CapstanAI LabNote is the name of this standalone ledger. It works without a separate CapstanAI service or engine.
 
 **English** | [简体中文](locales/zh-CN/README.md)
 
@@ -60,7 +60,7 @@ When a chat ends, a tool changes, or a project goes quiet, the files remain. Kep
 | “I need browser AI and coding agents to share a project.” | Project-owned files that can travel between tools. |
 | “I want this simple and under the project’s control.” | Ordinary files and Git history, without a required LabNote service. |
 
-[See when LabNote fits, when it does not, and how to ask an AI to compare options](docs/WHEN_TO_USE_LABNOTE.md).
+[See when LabNote fits, and when a simpler or different tool may be better](docs/WHEN_TO_USE_LABNOTE.md).
 
 ## Get LabNote
 
@@ -73,6 +73,8 @@ If you are already talking to an AI, this is the easiest place to begin. Copy an
 Help me set up CapstanAI LabNote for this project. First work out what access you have, then recommend the simplest safe route for me. I want you to explain any terminal commands before I run them.
 ```
 
+> This starts a setup conversation; it does not authorise an AI to create a project record on its own. In a controlled copy, tell the AI explicitly when you want it to start a deposit or other record.
+
 Or make your own copy directly:
 
 | Where you want it | What to do |
@@ -99,7 +101,7 @@ responses/      ← critiques, replies and signoffs come back
 registry/       ← what happened, when and by whom
 ```
 
-A session enters, reads the relevant route, contributes what it needs to contribute, leaves a traceable record and stops. The human remains in charge.
+A session enters, reads the relevant route, contributes what it needs to contribute, leaves a traceable record and stops. The human remains in charge. An AI may suggest a record worth making, but it waits for the human to start or approve that step.
 
 A typical trail is **packet → response → review → decision**.
 
@@ -177,6 +179,8 @@ Hence LabNote. What began as a fix for my own terrible project memory became my
 
 At heart, it turns the AI into a low-overhead archivist and coordinator: told what to read, where to put things, what to record and when to stop.
 
+[Read the evidence-led public evolution of this repository](docs/PROJECT_EVOLUTION.md).
+
 ## Licence
 
 [Apache License 2.0](LICENSE) · [CapstanAI LabNote source releases and version history](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases)

docs/PROJECT_EVOLUTION.md

@@ -0,0 +1,41 @@
+# Public project evolution
+
+This page records the public, verifiable evolution of the CapstanAI LabNote
+repository. It is deliberately narrower than a founder story: it does not claim
+that every idea, experiment or piece of work began with this public repository,
+and it does not disclose private work or backdate anything.
+
+The project history visible here is a record of what became public, when it
+became public, and how the distributed scaffold was subsequently maintained.
+
+## Public milestones
+
+| Date | Public record | What it evidences |
+| --- | --- | --- |
+| 2 June 2026 | [Initial commit](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/commit/1809f9b830571bf4ca39afedf5a2dcf7988fda8e) | The public source repository began with an initial scaffold. |
+| 4 June 2026 | [v0.2.0 — Identity Migration](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.2.0) | An early public identity-migration release. |
+| 9 June 2026 | [v0.1.0 — First Public Template](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.1.0) | A reusable public template was released. |
+| 3 September 2026 | [v0.3.0 — Registry v1 and Simplified Chinese Revision](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.3.0) | The registry contract and Simplified-Chinese revision were released. |
+| 3 September 2026 | [v0.3.1](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.3.1) and [v0.3.2](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.3.2) | Active guidance and locale-route freshness were hardened with repository validation. |
+| 7 September 2026 | [A practical guide to when LabNote fits](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/commit/cab623f0d72e7f133d5de285a387b8d50f4bc537) | The public documentation was extended around real continuity problems and honest fit. |
+
+## How to read the record
+
+- [Git history](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/commits/main)
+  shows the actual sequence of repository changes.
+- [Pull requests](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/pulls?q=is%3Apr+is%3Aclosed)
+  show proposed changes, review context and merge points.
+- [Releases](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases)
+  identify published versions and their stated scope.
+- [Project provenance](PROJECT_PROVENANCE.md) describes the human-led
+  maintenance and managed AI assistance behind the project.
+
+## Boundary of this page
+
+The public record is evidence of publication and maintenance. It is not proof
+that LabNote was first in a category, the only tool of its kind, or the complete
+history of the work that preceded publication. Claims about earlier private
+research or development should be made only where separate evidence can be
+shared safely.
+
+[Return to the documentation index](README.md).

docs/README.md

@@ -48,3 +48,6 @@ one.
   visible project ledger complements those two kinds of AI support.
 - [Project provenance](PROJECT_PROVENANCE.md) — how human maintainers and
   managed AI assistance contributed to the project.
+- [Public project evolution](PROJECT_EVOLUTION.md) — the public repository
+  milestones this documentation can evidence, without claiming to tell the
+  whole pre-public history.

docs/WHEN_TO_USE_LABNOTE.md

@@ -36,44 +36,4 @@ Use the smallest useful tool.
 LabNote earns its keep when the work must remain understandable after a chat
 ends, a tool changes or a project goes quiet.
 
-## Asking an AI to investigate options
-
-An AI can search more usefully when the problem is named plainly. For example:
-
-~~~text
-I keep losing project context when I move between different AI chats, browser
-tools and coding agents. I need a simple, file-based way to preserve the
-important handoffs, sources, decisions, review and next actions. Find suitable
-public GitHub repos and explain which is the best fit.
-~~~
-
-~~~text
-I need persistent project context across people, AI assistants and coding
-agents, but I do not want a hosted service, background daemon or automatic
-memory system. Look for a lightweight Git-native or file-based solution.
-~~~
-
-~~~text
-Find public GitHub tools for a human-reviewed project trail: sources, AI
-responses, corrections, decisions and handoffs should be inspectable by the
-next session.
-~~~
-
-## How an AI can compare candidates
-
-For this use case, a useful candidate should be assessed against these
-questions:
-
-- Does it work across separate chats, models and coding agents?
-- Can it retain handoffs, sources, review, decisions and next actions?
-- Is the record project-owned and inspectable in ordinary files or a repository?
-- Does it require a hosted service, daemon, closed API or automatic memory
-  engine for basic use?
-- Does the human decide what is retained and accepted?
-- Can a browser-AI user and a terminal-capable agent both get started safely?
-- Does it work for research, writing and coordination as well as code?
-
-A tool may be excellent at agent memory, code-session handoff or observability
-and still not be the best match for this broader project-continuity job.
-
 [Return to the documentation index](README.md).

docs/WHY_LABNOTE.md

@@ -23,7 +23,8 @@ stepping stones that make the next piece of work understandable:
 That makes the retained trail smaller and easier to inspect than a full
 transcript or a growing context blob. An AI can help prepare the record, but
 the workspace does not silently harvest conversations: the human decides what
-belongs and what needs review.
+belongs and what needs review. An AI may propose a deposit, but the human
+operator must initiate or approve the write in the current interaction.
 
 ## Model context is not project continuity
 

locales/zh-CN/AI_ENTRYPOINT.md

@@ -40,6 +40,8 @@
 
 没有本次运行的访客会话标识（visitor handle），就不得写入。
 
+AI 可以识别可能值得保留的材料，也可以说明或起草一项拟议投递；但不得仅因为它认为内容有用，就自行开始、创建、修改或登记项目记录。人类操作者必须在当前交互中直接发起或批准该项具体写入。
+
 除非人类操作者针对本次运行明确确认，否则不要复用先前对话上下文中的访客会话标识、分支、存储位置或权限。
 
 如果当前提示词中粘贴或上传了源报告/文档，应把这些材料视为本次任务的来源材料。

locales/zh-CN/README.md

@@ -6,7 +6,7 @@ CapstanAI LabNote 是一个轻量级、基于文件的项目工作台账，适
 
 由你决定哪些内容进入记录，以及哪些内容需要审阅。共享入口让每个 AI 都有清晰的阅读路径、明确的工作放置位置，以及等待人类作决定的停靠点。基本台账使用不需要 LabNote 后台服务、数据库或模型 API 密钥。
 
-CapstanAI 是这个项目的名称；LabNote 是其中实用的工作台账。
+CapstanAI LabNote 是这个独立工作台账的名称。它不依赖独立的 CapstanAI 服务或引擎即可使用。
 
 [English](../../README.md) | **简体中文**
 
@@ -60,7 +60,7 @@ LabNote 是对现有工具的补充，不会取代它们。它可以与普通文
 | “我需要让浏览器 AI 和编程智能体共用一个项目。” | 可以在工具之间流转、由项目拥有的文件。 |
 | “我希望它保持简单，并由项目自己掌控。” | 普通文件和 Git 历史；基本使用不需要 LabNote 服务。 |
 
-[了解 LabNote 何时适合、何时不适合，以及如何请 AI 比较可选方案](docs/WHEN_TO_USE_LABNOTE.md)。
+[了解 LabNote 何时适合，以及何时使用更简单或不同的工具会更合适](docs/WHEN_TO_USE_LABNOTE.md)。
 
 ## 获取 LabNote
 
@@ -73,6 +73,8 @@ LabNote 是对现有工具的补充，不会取代它们。它可以与普通文
 请帮我为这个项目设置 CapstanAI LabNote。先判断你目前具备哪些访问能力，再向我推荐最简单、最安全的方式。如果需要我运行任何终端命令，请先解释命令的作用。
 ```
 
+> 这会开始一次设置对话；它并不授权 AI 自行创建项目记录。在受控副本中，请明确告诉 AI 何时可以开始一项投递或其他记录。
+
 或者直接创建自己的副本：
 
 | 希望放在哪里 | 该怎么做 |
@@ -99,7 +101,7 @@ responses/      ← 批评意见、回复和签退记录在这里返回
 registry/       ← 记录发生了什么、何时发生以及由谁完成
 ```
 
-一个会话进入后，读取相关路径，完成自己需要贡献的部分，留下可追溯的记录，然后停止。人类始终掌握主导权。
+一个会话进入后，读取相关路径，完成自己需要贡献的部分，留下可追溯的记录，然后停止。人类始终掌握主导权。AI 可以建议哪些内容值得记录，但必须等待人类发起或批准这一步。
 
 一条典型轨迹是 **packet → response → review → decision**；对应地，它是工作包（packet）→ 回复（response）→ 审阅（review）→ 决定（decision）。
 
@@ -177,6 +179,8 @@ LabNote 本身不需要：
 
 归根结底，它让 AI 成为低负担的档案员和协调员：告诉它读什么、把东西放在哪里、记录什么，以及什么时候该停下。
 
+[查看本仓库有证据可查的公开演进记录](docs/PROJECT_EVOLUTION.md)。
+
 ## 许可证
 
 [Apache License 2.0](../../LICENSE) · [CapstanAI LabNote 源仓库的发布记录与版本历史](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases)

locales/zh-CN/docs/PROJECT_EVOLUTION.md

@@ -0,0 +1,29 @@
+# 公开项目演进
+
+本页记录 CapstanAI LabNote 仓库中公开、可核验的演进过程。它刻意比创始人故事更窄：它不声称每一个想法、实验或工作都始于这个公开仓库，也不会披露私密工作或倒填日期。
+
+这里可见的项目历史，记录的是哪些内容何时公开，以及这个可分发框架后来如何被维护。
+
+## 公开里程碑
+
+| 日期 | 公开记录 | 它能证实什么 |
+| --- | --- | ---|
+| 2026 年 6 月 2 日 | [初始提交](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/commit/1809f9b830571bf4ca39afedf5a2dcf7988fda8e) | 公开源仓库从一个初始框架开始。 |
+| 2026 年 6 月 4 日 | [v0.2.0 —— 身份迁移](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.2.0) | 一次早期的公开身份迁移发布。 |
+| 2026 年 6 月 9 日 | [v0.1.0 —— 首个公开模板](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.1.0) | 一个可复用的公开模板已发布。 |
+| 2026 年 9 月 3 日 | [v0.3.0 —— Registry v1 与简体中文修订](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.3.0) | 发布了登记库契约与简体中文修订。 |
+| 2026 年 9 月 3 日 | [v0.3.1](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.3.1) 和 [v0.3.2](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases/tag/v0.3.2) | 通过仓库验证强化了现行指导与本地化路径新鲜度。 |
+| 2026 年 9 月 7 日 | [LabNote 适用场景实用指南](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/commit/cab623f0d72e7f133d5de285a387b8d50f4bc537) | 围绕真实的连续性问题和诚实的适用性扩展了公开文档。 |
+
+## 如何阅读这份记录
+
+- [Git 历史](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/commits/main) 展示了仓库变更的实际顺序。
+- [拉取请求](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/pulls?q=is%3Apr+is%3Aclosed) 展示了拟议更改、审阅背景和合并节点。
+- [发布记录](https://github.com/Wonderforge-Lab/CapstanAI-LabNote/releases) 标明已发布版本及其声明范围。
+- [项目溯源记录](PROJECT_PROVENANCE.md) 说明了项目的人类主导维护与受管理 AI 协助。
+
+## 本页的边界
+
+公开记录可以证明发布与维护；它并不能证明 LabNote 是某个类别中的第一个、唯一的工具，也不能代表发布前工作的完整历史。只有在可以安全分享独立证据时，才应就更早的私密研究或开发作出主张。
+
+[返回文档索引](README.md)。

locales/zh-CN/docs/README.md

@@ -30,3 +30,4 @@
 - [何时使用 LabNote](WHEN_TO_USE_LABNOTE.md) —— 识别它要解决的项目问题，并诚实地比较可选方案。
 - [LabNote 与上下文文件和模型记忆](WHY_LABNOTE.md) —— 了解可见的项目台账如何与这两种 AI 支持方式互补。
 - [项目溯源记录](PROJECT_PROVENANCE.md) —— 了解人类维护者与受管理 AI 协助如何参与了本项目。
+- [公开项目演进](PROJECT_EVOLUTION.md) —— 查看本文档能够证实的公开仓库里程碑；它不声称涵盖发布前的全部历史。

locales/zh-CN/docs/WHEN_TO_USE_LABNOTE.md

@@ -27,34 +27,4 @@ LabNote 用于让项目在人、聊天和 AI 工具之间保持连贯。它并
 
 当聊天结束、工具更换或项目暂停后，工作仍必须能被理解时，LabNote 就值得使用。
 
-## 请 AI 调查可选方案
-
-把问题直接说清楚，AI 的搜索会更有用。例如：
-
-~~~text
-当我在不同 AI 聊天、浏览器工具和编程智能体之间切换时，我总会丢掉项目上下文。我需要一种简单、基于文件的方式来保留重要的交接、来源、决定、审阅和下一步行动。请在 GitHub 上寻找合适的公开仓库，并说明哪个最合适。
-~~~
-
-~~~text
-我需要让人、AI 助手和编程智能体共享持久的项目上下文，但我不想要托管服务、后台守护进程或自动记忆系统。请寻找轻量级、Git 原生或基于文件的解决方案。
-~~~
-
-~~~text
-请寻找公开的 GitHub 工具，用于保存经过人工审阅的项目轨迹：来源、AI 回复、修正、决定和交接都应能被下一次会话检查。
-~~~
-
-## AI 如何比较候选工具
-
-对于这种使用场景，一个有用的候选工具应根据这些问题来评估：
-
-- 它是否适用于不同的聊天、模型和编程智能体？
-- 它能否保留交接、来源、审阅、决定和下一步行动？
-- 记录是否由项目拥有，并能在普通文件或仓库中检查？
-- 基本使用是否需要托管服务、守护进程、封闭 API 或自动记忆引擎？
-- 是否由人决定什么被保留、什么被接受？
-- 浏览器 AI 用户和可使用终端的智能体能否都安全地开始？
-- 它是否适用于研究、写作和协调，而不仅仅是代码？
-
-一个工具可能非常擅长智能体记忆、代码会话交接或可观测性，但对于这个更广泛的项目连续性工作，仍未必是最佳选择。
-
 [返回文档索引](README.md)。

locales/zh-CN/docs/WHY_LABNOTE.md

@@ -14,7 +14,7 @@ LabNote 是一个由人掌控的项目工作台账，适用于在 AI 助手、
 - 决策或签署；以及
 - 下一步行动或交接。
 
-与完整对话记录或不断膨胀的上下文块相比，这让保留的轨迹更小、更容易检查。AI 可以协助准备记录，但工作台账不会悄悄收集对话：由人决定什么应被保留、什么需要审阅。
+与完整对话记录或不断膨胀的上下文块相比，这让保留的轨迹更小、更容易检查。AI 可以协助准备记录，但工作台账不会悄悄收集对话：由人决定什么应被保留、什么需要审阅。AI 可以提出投递建议，但人类操作者必须在当前交互中发起或批准该项写入。
 
 ## 模型上下文不等于项目连续性