From b37d260a95746e1895d5409e69096a4456948051 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E9=99=B3=E6=AC=A3=E6=80=A1?= Date: Thu, 26 Mar 2026 09:23:26 +0800 Subject: [PATCH 1/4] docs: update logo to social-preview and normalize product name to Label Suite - Replace icon-colored.svg with social-preview.svg in README.md and README.zh-TW.md - Normalize product name from LabelSuite to Label Suite across all docs and assets - Add aria-label i18n support (data-i18n-aria-label) to login and dashboard prototypes Co-Authored-By: Claude Sonnet 4.6 --- .claude/agents/nlp-research-advisor.md | 2 +- .specify/memory/lessons-learned.md | 4 ++-- README.md | 12 +++++----- README.zh-TW.md | 12 +++++----- assets/logo/README.md | 12 +++++----- assets/logo/icon-colored.svg | 2 +- assets/logo/icon.svg | 2 +- assets/logo/logo-horizontal.svg | 2 +- design-system/labelsuite/MASTER.md | 2 +- .../thesis/thesis-ch1-introduction.d2 | 2 +- docs/research/interviews/README.md | 2 +- docs/research/tool-analysis.md | 24 +++++++++---------- docs/thesis/outline.md | 12 +++++----- docs/thesis/outline.zh-TW.md | 14 +++++------ prototype/index.html | 2 +- prototype/pages/dashboard.html | 12 +++++----- prototype/pages/login.html | 10 ++++---- specs/001-sso-login/spec.md | 2 +- 18 files changed, 65 insertions(+), 65 deletions(-) diff --git a/.claude/agents/nlp-research-advisor.md b/.claude/agents/nlp-research-advisor.md index b044b7f..55e581c 100644 --- a/.claude/agents/nlp-research-advisor.md +++ b/.claude/agents/nlp-research-advisor.md @@ -19,7 +19,7 @@ You are an NLP research advisor with deep expertise in Chinese NLP, data annotat ## Project Context Academic background for this project: -- **System Name**: LabelSuite +- **System Name**: Label Suite - **Advisor**: Professor Lung-Hao Lee, Natural Language Processing Laboratory - **Paper Type**: Demo Paper (system/tool paper) - **Core Contribution**: Config-driven general-purpose NLP annotation platform with integrated annotator management diff --git a/.specify/memory/lessons-learned.md b/.specify/memory/lessons-learned.md index 980a3a1..14019de 100644 --- a/.specify/memory/lessons-learned.md +++ b/.specify/memory/lessons-learned.md @@ -19,7 +19,7 @@ Claude Code directly committed and pushed logo assets to the `main` branch witho **Incorrect workflow executed:** ```bash git add assets/ .gitignore README.md README.zh-TW.md -git commit -m "feat: add LabelSuite logo and brand assets" +git commit -m "feat: add Label Suite logo and brand assets" git push origin main # ❌ VIOLATION ``` @@ -51,7 +51,7 @@ git push origin main # ❌ VIOLATION git checkout -b feat/logo-and-brand-assets # Re-created all assets git add assets/ .gitignore README.md README.zh-TW.md - git commit -m "feat: add LabelSuite logo and brand assets" + git commit -m "feat: add Label Suite logo and brand assets" git push -u origin feat/logo-and-brand-assets ``` diff --git a/README.md b/README.md index 1c76a46..2677664 100644 --- a/README.md +++ b/README.md @@ -1,8 +1,8 @@

- LabelSuite Logo + Label Suite Logo

-

LabelSuite

+

Label Suite

繁體中文 | English @@ -23,7 +23,7 @@ Existing annotation platforms such as [Label Studio](https://labelstud.io/) are - **Fragmented workflows:** Task configuration, labeling, and dataset analysis are often handled by separate tools or ad-hoc scripts, forcing researchers to repeatedly build one-off systems from scratch. - **No dataset quality visibility:** Existing tools provide no built-in dataset statistics, forcing researchers to write analysis scripts after each labeling round. -**LabelSuite** aims to eliminate these pain points by providing a lightweight, config-driven annotation platform that any NLP research team can launch with minimal setup. +**Label Suite** aims to eliminate these pain points by providing a lightweight, config-driven annotation platform that any NLP research team can launch with minimal setup. --- @@ -85,7 +85,7 @@ This project is positioned as a **Demo Paper**, with its core value in: ## Comparison with Label Studio -| Feature | Label Studio | **LabelSuite** | +| Feature | Label Studio | **Label Suite** | |---|---|---| | Easy setup (no server config) | ✗ | ✓ | | Config-driven task definition | Partial | ✓ | @@ -115,7 +115,7 @@ This project is positioned as a **Demo Paper**, with its core value in: } }}%% gantt - title LabelSuite — Research Roadmap + title Label Suite — Research Roadmap dateFormat YYYY-MM-DD axisFormat %b %Y @@ -152,7 +152,7 @@ gantt - [ ] Survey Label Studio and identify pain points in setup, usability, and annotator management - [ ] Conduct UX interviews and distribute a pain-point questionnaire to target users (researchers, annotators) - [ ] Survey related academic papers on annotation platforms to establish positioning for the Related Work section -- [ ] Define the system's contribution: clarify how LabelSuite is simpler and more usable than Label Studio +- [ ] Define the system's contribution: clarify how Label Suite is simpler and more usable than Label Studio - [ ] Study Demo Paper examples from target venue proceedings to understand structure, length, and demonstration requirements ### Phase 2 — System Design & General-Purpose Architecture (Month 5–8) diff --git a/README.zh-TW.md b/README.zh-TW.md index 9a90b0a..8478bce 100644 --- a/README.zh-TW.md +++ b/README.zh-TW.md @@ -1,8 +1,8 @@

- LabelSuite Logo + Label Suite Logo

-

LabelSuite

+

Label Suite

繁體中文 | English @@ -23,7 +23,7 @@ - **工作流程分散：** 任務配置、標記執行與資料集分析往往需要不同工具或自行撰寫腳本，導致研究團隊重複開發一次性系統。 - **缺乏資料集品質可視性：** 現有工具未提供內建的資料集統計功能，研究人員每次標記後都需撰寫分析腳本才能了解資料品質。 -**LabelSuite** 旨在消除上述痛點，提供一個輕量、配置驅動的標記平台，讓任何 NLP 研究團隊都能以最少的設定快速啟動。 +**Label Suite** 旨在消除上述痛點，提供一個輕量、配置驅動的標記平台，讓任何 NLP 研究團隊都能以最少的設定快速啟動。 --- @@ -85,7 +85,7 @@ ## 與 Label Studio 比較 -| 功能 | Label Studio | **LabelSuite** | +| 功能 | Label Studio | **Label Suite** | |---|---|---| | 簡易架設（免伺服器配置） | ✗ | ✓ | | Config 驅動的任務定義 | 部分支援 | ✓ | @@ -115,7 +115,7 @@ } }}%% gantt - title LabelSuite — 研究規劃 + title Label Suite — 研究規劃 dateFormat YYYY-MM-DD axisFormat %b %Y @@ -152,7 +152,7 @@ gantt - [ ] 調研 Label Studio，找出架設、易用性與標記員管理上的痛點 - [ ] 進行 UX 需求訪談，並發放使用者痛點問卷（目標對象：研究人員、標記人員） - [ ] 調查標記平台的相關學術論文，為 Related Work 建立定位基礎 -- [ ] 確立論文貢獻點：定義 LabelSuite 如何比 Label Studio 更好用、更簡便 +- [ ] 確立論文貢獻點：定義 Label Suite 如何比 Label Studio 更好用、更簡便 - [ ] 研讀目標投稿場合的 Demo Paper 範例，了解結構、篇幅與展示需求 ### 第二階段 — 系統設計與一般化架構（第 5–8 個月） diff --git a/assets/logo/README.md b/assets/logo/README.md index e5fc7cb..72d75b4 100644 --- a/assets/logo/README.md +++ b/assets/logo/README.md @@ -1,6 +1,6 @@ -# LabelSuite Logo & Brand Assets +# Label Suite Logo & Brand Assets -This directory contains the official LabelSuite logo and brand assets. +This directory contains the official Label Suite logo and brand assets. ## Files @@ -10,7 +10,7 @@ This directory contains the official LabelSuite logo and brand assets. |------|-------------|-------| | `icon.svg` | Tag icon only (monochrome) | Favicon, small icons | | `icon-colored.svg` | Icon with primary color background (48×48) | App icon, profile picture | -| `logo-horizontal.svg` | Icon + "LabelSuite" text | Header, documentation | +| `logo-horizontal.svg` | Icon + "Label Suite" text | Header, documentation | | `social-preview.svg` | GitHub social preview (1280×640) | GitHub repository preview | ### PNG (Raster) @@ -101,10 +101,10 @@ GitHub repositories don't have custom icons by default, but you can: ```markdown

- LabelSuite Logo + Label Suite Logo

-

LabelSuite

+

Label Suite

NLP Annotation Portal

``` @@ -132,4 +132,4 @@ Then add to your HTML: ## License -These assets are part of the LabelSuite project. See the project's main LICENSE file for details. +These assets are part of the Label Suite project. See the project's main LICENSE file for details. diff --git a/assets/logo/icon-colored.svg b/assets/logo/icon-colored.svg index d8b393e..430e08e 100644 --- a/assets/logo/icon-colored.svg +++ b/assets/logo/icon-colored.svg @@ -1,4 +1,4 @@ - + diff --git a/assets/logo/icon.svg b/assets/logo/icon.svg index 3d90c36..23d2650 100644 --- a/assets/logo/icon.svg +++ b/assets/logo/icon.svg @@ -1,4 +1,4 @@ - + + diff --git a/design-system/labelsuite/MASTER.md b/design-system/labelsuite/MASTER.md index 1d8af62..95dc506 100644 --- a/design-system/labelsuite/MASTER.md +++ b/design-system/labelsuite/MASTER.md @@ -6,7 +6,7 @@ --- -**Project:** LabelSuite +**Project:** Label Suite **Generated:** 2026-03-25 11:49:50 **Category:** Micro SaaS diff --git a/docs/diagrams/thesis/thesis-ch1-introduction.d2 b/docs/diagrams/thesis/thesis-ch1-introduction.d2 index 50a202d..9c6c69d 100644 --- a/docs/diagrams/thesis/thesis-ch1-introduction.d2 +++ b/docs/diagrams/thesis/thesis-ch1-introduction.d2 @@ -31,7 +31,7 @@ pain_points: Pain Points { p5: No Dataset Quality Visibility {class: base} } -portal: LabelSuite { +portal: Label Suite { class: module o1: Config-driven Launch {class: base} diff --git a/docs/research/interviews/README.md b/docs/research/interviews/README.md index 860a3a2..b7e0347 100644 --- a/docs/research/interviews/README.md +++ b/docs/research/interviews/README.md @@ -1,4 +1,4 @@ -# LabelSuite 使用者需求訪談 +# Label Suite 使用者需求訪談 ## 訪談目的 diff --git a/docs/research/tool-analysis.md b/docs/research/tool-analysis.md index 7b760d8..cb05e6a 100644 --- a/docs/research/tool-analysis.md +++ b/docs/research/tool-analysis.md @@ -1,12 +1,12 @@ # 標記工具分析 -> 針對 Label Studio（參考基準）與 LabelSuite 進行多維度比較，涵蓋部署方式、任務支援、標記員管理、資料集統計、使用者體驗與技術架構，用於定義 LabelSuite 在 Demo Paper 中的貢獻與差異化定位。 +> 針對 Label Studio（參考基準）與 Label Suite 進行多維度比較，涵蓋部署方式、任務支援、標記員管理、資料集統計、使用者體驗與技術架構，用於定義 Label Suite 在 Demo Paper 中的貢獻與差異化定位。 --- ## 表一 — 系統概覽與定位 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **類型** | 通用型資料標記平台 | 具實驗室管理功能的 NLP 專用標記平台 | | **目標使用者** | 企業、ML 工程師、資料團隊 | 超級管理員、實驗室研究生、兼職工讀生標記員 | @@ -19,7 +19,7 @@ ## 表二 — 部署與設定 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **設定方式** | 手動伺服器設定（Docker、環境變數、DB 遷移） | 單一設定檔（YAML/JSON）+ 一鍵啟動 | | **工程背景需求** | 需要 DevOps / 工程背景 | 無需工程背景 | @@ -32,7 +32,7 @@ ## 表三 — NLP 任務類型支援 -| 任務類型 | Label Studio | LabelSuite | +| 任務類型 | Label Studio | Label Suite | |---|---|---| | **單句任務**（分類 / 評分） | ✅（文字分類範本） | ✅（原生 NLP 範本） | | **句對任務**（相似度 / 蘊含） | 🟡（需手動設定） | ✅（專用範本） | @@ -45,7 +45,7 @@ ## 表四 — 標記員管理 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **帳號管理** | 基本使用者帳號（無標記員角色區分） | 完整 CRUD，角色型存取控制（Admin / Researcher / Annotator） | | **工讀生支援** | ❌ | ✅（專為實驗室工讀生工作流程設計） | @@ -58,7 +58,7 @@ ## 表五 — 資料集統計分析 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **內建統計功能** | ❌（無原生統計儀表板） | ✅（#Sentence、#Token、#Label） | | **標籤分布視圖** | ❌ | ✅（即時類別不均衡偵測） | @@ -71,7 +71,7 @@ ## 表六 — 標記工作流程 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **試跑模式（Dry Run）** | ❌ | ✅（正式收集前驗證設定） | | **正式模式（Official Run）** | ❌（單一模式） | ✅（資料與試跑模式嚴格隔離） | @@ -84,7 +84,7 @@ ## 表七 — 介面與使用者體驗 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **介面複雜度** | 高（功能豐富，學習曲線陡峭） | 低（精簡，聚焦任務） | | **非工程師入門** | 困難（需閱讀文件） | 簡單（由設定檔結構引導） | @@ -97,7 +97,7 @@ ## 表八 — 技術架構 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **後端框架** | Django（Python）— 重量級，同步優先 | FastAPI（Python）— 輕量級，非同步原生 | | **前端框架** | React 18 + TypeScript（Webpack 5、Nx monorepo） | React 18 + TypeScript + Vite | @@ -113,7 +113,7 @@ ## 表九 — 研究與學術適配性 -| 比較維度 | Label Studio | LabelSuite | +| 比較維度 | Label Studio | Label Suite | |---|---|---| | **學術實驗室導向** | ❌（以企業 / ML 團隊為主） | ✅（專為 NLP 研究實驗室設計） | | **中文 NLP 優化** | ❌ | ✅（醫療、情感、心理領域） | @@ -124,9 +124,9 @@ --- -## 總結 — LabelSuite 差異化定位 +## 總結 — Label Suite 差異化定位 -LabelSuite 鎖定 Label Studio 在**學術 NLP 實驗室**場景中的不足，提供： +Label Suite 鎖定 Label Studio 在**學術 NLP 實驗室**場景中的不足，提供： 1. **快速、零工程部署** — 設定檔取代繁瑣的伺服器設定流程 2. **標記員生命週期管理** — 所有現有工具均缺乏的內建人資功能 diff --git a/docs/thesis/outline.md b/docs/thesis/outline.md index 3876b34..ae294eb 100644 --- a/docs/thesis/outline.md +++ b/docs/thesis/outline.md @@ -1,6 +1,6 @@ -# Thesis Outline — LabelSuite +# Thesis Outline — Label Suite -**Title (tentative):** LabelSuite: A Config-Driven NLP Annotation Platform with Integrated Annotator Management +**Title (tentative):** Label Suite: A Config-Driven NLP Annotation Platform with Integrated Annotator Management **Type:** Demo Paper (Master's Thesis) @@ -26,7 +26,7 @@ ### 1.3 Research Objectives -- **Developing a General-Purpose Annotation Platform:** Build a lightweight platform named **`LabelSuite`** targeting academic NLP labs, enabling rapid deployment of diverse annotation tasks through simple configuration files. +- **Developing a General-Purpose Annotation Platform:** Build a lightweight platform named **`Label Suite`** targeting academic NLP labs, enabling rapid deployment of diverse annotation tasks through simple configuration files. - **Config-driven Launch:** Enable rapid deployment of labeling servers through simple config files, replacing traditional complex system development workflows. - **Annotator Lifecycle Management:** Provide end-to-end annotator management covering account administration, working hours tracking, and salary estimation to streamline lab operations. - **Dataset Quality Visibility:** Provide real-time dataset statistics (#Sentence, #Token, #Label) to help researchers inspect data characteristics and annotation quality without external scripts. @@ -52,7 +52,7 @@ ### 2.2 Survey of Labeling Platforms & Tools - **Label Studio Analysis:** An open-source data annotation platform (Apache 2.0) adopted by enterprises such as NVIDIA, Meta, and IBM. It excels in multi-modal support (image, audio, text, video, time series) and LLM fine-tuning data preparation. However, its server setup is tedious for non-engineering researchers, and it lacks built-in dataset analytics and annotator management features. -- **Differentiated Positioning of LabelSuite:** Taking Label Studio as the reference baseline, `LabelSuite` targets academic NLP labs and differentiates by: (1) enabling config-driven rapid deployment without engineering overhead; (2) adding annotator lifecycle management (account administration, working hours tracking, salary estimation) absent in Label Studio; (3) integrating built-in dataset statistics (#Sentence, #Token, #Label) for real-time quality monitoring. +- **Differentiated Positioning of Label Suite:** Taking Label Studio as the reference baseline, `Label Suite` targets academic NLP labs and differentiates by: (1) enabling config-driven rapid deployment without engineering overhead; (2) adding annotator lifecycle management (account administration, working hours tracking, salary estimation) absent in Label Studio; (3) integrating built-in dataset statistics (#Sentence, #Token, #Label) for real-time quality monitoring. ### 2.3 Survey of Current Workflow Pain Points @@ -103,7 +103,7 @@ ### 4.2 System Interface Showcase and UX Optimization -- **Intuitive Design:** Addresses the pain points of Label Studio's complex and cluttered setup, demonstrating how **`LabelSuite`** simplifies task management and labeling workflows. +- **Intuitive Design:** Addresses the pain points of Label Studio's complex and cluttered setup, demonstrating how **`Label Suite`** simplifies task management and labeling workflows. - **Visual Comparison:** Screenshot-based comparison of this system vs Label Studio for the same labeling task, highlighting the "rapid launch," "annotator management," and "built-in analytics" advantages. ### 4.3 Implementation of Core Modules @@ -145,7 +145,7 @@ ### 6.1 Conclusion -- **Summary of Contributions:** Summarizes how **`LabelSuite`** addresses the practical pain points of workflow fragmentation, "reinventing the wheel," annotator management overhead, and the absence of built-in dataset quality visibility in academic NLP labs. +- **Summary of Contributions:** Summarizes how **`Label Suite`** addresses the practical pain points of workflow fragmentation, "reinventing the wheel," annotator management overhead, and the absence of built-in dataset quality visibility in academic NLP labs. - **Realization of Demo Paper Value:** Emphasizes the system's positioning as a config-driven, annotator-centered alternative to Label Studio for academic research teams, with open-source reuse value. ### 6.2 Research Limitations (New Section) diff --git a/docs/thesis/outline.zh-TW.md b/docs/thesis/outline.zh-TW.md index 0dfebd6..5e16f23 100644 --- a/docs/thesis/outline.zh-TW.md +++ b/docs/thesis/outline.zh-TW.md @@ -1,8 +1,8 @@ -# 論文大綱 — LabelSuite +# 論文大綱 — Label Suite -**論文名稱（中文）：** LabelSuite：具整合式標記員管理之配置驅動 NLP 標記平台設計與實作 +**論文名稱（中文）：** Label Suite：具整合式標記員管理之配置驅動 NLP 標記平台設計與實作 -**論文名稱（英文）：** LabelSuite：A Config-Driven NLP Annotation Platform with Integrated Annotator Management +**論文名稱（英文）：** Label Suite：A Config-Driven NLP Annotation Platform with Integrated Annotator Management **類型：** Demo Paper（碩士論文） @@ -32,7 +32,7 @@ ### 1.3 研究目的 (Research Objectives) -- **開發通用型標記平台 (Developing a General-Purpose Annotation Platform)：** 建立一個名為 **`LabelSuite`** 的輕量化平台，以學術 NLP 實驗室為目標，透過簡單設定檔即可快速部署多様化標記任務。 +- **開發通用型標記平台 (Developing a General-Purpose Annotation Platform)：** 建立一個名為 **`Label Suite`** 的輕量化平台，以學術 NLP 實驗室為目標，透過簡單設定檔即可快速部署多様化標記任務。 - **配置化啟動 (Config-driven Launch)：** 實現透過簡單設定檔 (Config file) 即可快速啟動標記伺服器的機制，取代傳統繁雜的系統開發流程。 - **標記員全生命週期管理 (Annotator Lifecycle Management)：** 提供涵蓋帳號管理、工時追蹤與薪資試算的端到端標記員管理功能，簡化實驗室行政作業。 - **資料集品質可視性 (Dataset Quality Visibility)：** 提供即時資料集統計指標（#Sentence、#Token、#Label），協助研究人員無需外部腳本即可快速掌握資料集特性與標記品質。 @@ -58,7 +58,7 @@ ### 2.2 標記平台與工具調研 (Survey of Labeling Platforms & Tools) - **Label Studio 分析 (Label Studio Analysis)：** 以 Apache 2.0 授權開源，獲 NVIDIA、Meta、IBM 等企業採用。核心強項為多模態支援（圖像、音頻、文本、影片）與 LLM 微調資料準備。然而伺服器架設對非工程背景研究員過於繁瑣，且不具備標記員管理與內建資料集統計功能。 -- **LabelSuite 的差異化定位 (Differentiated Positioning)：** 以 Label Studio 為參考標的，`LabelSuite` 針對學術 NLP 實驗室進行差異化：(1) 配置驅動快速部署，無需工程背景；(2) 新增 Label Studio 所缺乏的標記員全生命週期管理（帳號管理、工時追蹤、薪資試算）；(3) 整合內建資料集統計（#Sentence、#Token、#Label）以支援即時品質監控。 +- **Label Suite 的差異化定位 (Differentiated Positioning)：** 以 Label Studio 為參考標的，`Label Suite` 針對學術 NLP 實驗室進行差異化：(1) 配置驅動快速部署，無需工程背景；(2) 新增 Label Studio 所缺乏的標記員全生命週期管理（帳號管理、工時追蹤、薪資試算）；(3) 整合內建資料集統計（#Sentence、#Token、#Label）以支援即時品質監控。 ### 2.3 盤點現有流程的痛點 (Survey of Current Workflow Pain Points) @@ -109,7 +109,7 @@ ### 4.2 系統介面展示與 UX 優化 (System Interface Showcase and UX Optimization) -- **直覺式設計 (Intuitive Design)：** 針對 Label Studio 介面設置繁瑣的痛點，展示 **`LabelSuite`** 如何簡化任務管理與標記流程。 +- **直覺式設計 (Intuitive Design)：** 針對 Label Studio 介面設置繁瑣的痛點，展示 **`Label Suite`** 如何簡化任務管理與標記流程。 - **可視化對比 (Visual Comparison)：** 以截圖形式對比本系統與 **Label Studio** 在執行同一標記任務時的步驟差異，強調本系統的「快速啟動」、「標記員管理」與「內建分析」優勢。 ### 4.3 核心模組實作 (Implementation of Core Modules) @@ -151,7 +151,7 @@ ### 6.1 研究結論 (Conclusion) -- **貢獻總結 (Summary of Contributions)：** 總結 **`LabelSuite`** 如何解決標記流程斷裂、「重複造輪子」、實驗室標記員管理負擔與缺乏內建資料集品質可視性等實務痛點。 +- **貢獻總結 (Summary of Contributions)：** 總結 **`Label Suite`** 如何解決標記流程斷裂、「重複造輪子」、實驗室標記員管理負擔與缺乏內建資料集品質可視性等實務痛點。 - **Demo Paper 價值體現 (Realization of Demo Paper Value)：** 強調系統作為學術研究團隊以配置驅動、以標記員為中心之 Label Studio 替代方案的開源重用價值。 ### 6.2 研究限制 (Research Limitations) (New Section) diff --git a/prototype/index.html b/prototype/index.html index 8fe7801..16cf4d2 100644 --- a/prototype/index.html +++ b/prototype/index.html @@ -3,7 +3,7 @@ - LabelSuite Prototype + Label Suite Prototype

Redirecting to login page…

diff --git a/prototype/pages/dashboard.html b/prototype/pages/dashboard.html index f405ae1..f14f54e 100644 --- a/prototype/pages/dashboard.html +++ b/prototype/pages/dashboard.html @@ -3,7 +3,7 @@ - 儀表板 — LabelSuite + 儀表板 — Label Suite @@ -40,14 +40,14 @@
- +
- LabelSuite + Label Suite @@ -337,7 +337,7 @@

- 登入 — LabelSuite + 登入 — Label Suite @@ -64,7 +64,7 @@

-

LabelSuite

+

Label Suite

NLP 標注平台

@@ -148,7 +148,7 @@

登入登入 Date: Thu, 26 Mar 2026 09:47:14 +0800 Subject: [PATCH 2/4] docs: record second main branch violation incident in lessons-learned Add 2026-03-26 incident entry documenting root cause (lessons-learned.md not in auto-memory path) and remediation (moved rule to auto-memory). Co-Authored-By: Claude Sonnet 4.6 --- .specify/memory/lessons-learned.md | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/.specify/memory/lessons-learned.md b/.specify/memory/lessons-learned.md index 14019de..c857031 100644 --- a/.specify/memory/lessons-learned.md +++ b/.specify/memory/lessons-learned.md @@ -126,6 +126,31 @@ Before executing `git push`: --- +## 2026-03-26: Direct Commit to Main Branch — Second Violation + +### Incident Summary +**Date:** 2026-03-26 +**Severity:** High +**Type:** Git Workflow Violation (repeat offense) +**Reporter:** User (mandychen) +**Agent:** Claude Code (Sonnet 4.6) + +### What Happened +Claude Code committed directly to `main` without creating a feature branch, despite the same violation being recorded on 2025-03-25. + +### Root Cause Analysis +`lessons-learned.md` is stored in `.specify/memory/` which is NOT part of Claude's auto-memory system. The file is never automatically loaded into context, so the lesson had no effect on subsequent conversations. + +### Remediation Steps Taken +1. Created branch `docs/logo-and-i18n-update` at current HEAD +2. Reset `main` to `origin/main` with `git reset --hard HEAD~1` +3. Added feedback memory to auto-memory system at `~/.claude/projects/.../memory/feedback_never_commit_to_main.md` + +### Prevention +Lesson moved to auto-memory so it loads automatically in every future conversation. + +--- + ## Template for Future Incidents ```markdown From 2db1b3ebdfd9daa903f0ae93b37536939e847e53 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E9=99=B3=E6=AC=A3=E6=80=A1?= Date: Thu, 26 Mar 2026 09:59:07 +0800 Subject: [PATCH 3/4] feat(assets): add banner.svg and update README to use banner logo MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - Add 1280×200 horizontal banner (icon + title + subtitle, white bg) - Update README.md and README.zh-TW.md to use banner.svg Co-Authored-By: Claude Sonnet 4.6 --- README.md | 2 +- README.zh-TW.md | 2 +- assets/logo/banner.svg | 30 ++++++++++++++++++++++++++++++ 3 files changed, 32 insertions(+), 2 deletions(-) create mode 100644 assets/logo/banner.svg diff --git a/README.md b/README.md index 2677664..39fc47e 100644 --- a/README.md +++ b/README.md @@ -1,5 +1,5 @@

- Label Suite Logo + Label Suite

Label Suite

diff --git a/README.zh-TW.md b/README.zh-TW.md index 8478bce..5230004 100644 --- a/README.zh-TW.md +++ b/README.zh-TW.md @@ -1,5 +1,5 @@

- Label Suite Logo + Label Suite

Label Suite

diff --git a/assets/logo/banner.svg b/assets/logo/banner.svg new file mode 100644 index 0000000..fcaa873 --- /dev/null +++ b/assets/logo/banner.svg @@ -0,0 +1,30 @@ + + + + + + + + + + + + + + + + + + Label Suite + + + NLP Annotation Portal + + From b42999fca466cbc36842a95994316318817f89db Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E9=99=B3=E6=AC=A3=E6=80=A1?= Date: Thu, 26 Mar 2026 10:45:58 +0800 Subject: [PATCH 4/4] fix: restore labelsuite-lang storage key and add aria-home i18n to dashboard - Fix finding 2: add data-i18n-aria-label="aria-home" to dashboard logo link with zh-TW/en translations so aria-label updates on language switch - Fix finding 3: revert STORAGE_KEY from 'Label Suite-lang' back to 'labelsuite-lang' to preserve cross-page language persistence Co-Authored-By: Claude Sonnet 4.6 --- prototype/pages/dashboard.html | 6 ++++-- prototype/pages/login.html | 2 +- 2 files changed, 5 insertions(+), 3 deletions(-) diff --git a/prototype/pages/dashboard.html b/prototype/pages/dashboard.html index f14f54e..32132c4 100644 --- a/prototype/pages/dashboard.html +++ b/prototype/pages/dashboard.html @@ -40,7 +40,7 @@
- +

登入