v1.3.3

Changelog

[v1.3.3] - 2026-04-18

Added

• Restored a safe backend/.env.example template with production-ready placeholders for GitHub, AI, OAuth, Redis, and database settings
• Added a clearer .env.production.example flow for Docker production deployments using localhost binding behind an HTTPS reverse proxy

Changed

• Production Docker Compose now binds the frontend to 127.0.0.1:8080 by default instead of exposing port 80 directly
• Deployment documentation now standardizes the recommended production topology: Docker on localhost, Nginx or Caddy in front, and explicit PUBLIC_BIND_IP control when direct exposure is intentional
• Backend and deployment docs now use container-aware defaults such as REDIS_URL=redis://redis:6379 and domain-scoped CORS_ORIGINS

[v1.3.2] - 2026-04-18

Added

• GitScore dimension explanations in the backend payload and AI analysis view, including evidence, next steps, low-data handling, and strongest/weakest dimension cues
• Structured repository analysis fields for strengths, risks, resume-ready bullets, next steps, showcase fit, confidence, and evidence-backed engineering signals
• Focused frontend regression coverage for repository analysis URL persistence and export PDF reuse flows

Changed

• Repository analysis prompts now incorporate README headings plus engineering-signal detection for tests, CI, docs, examples, container files, frontend/backend structure, and API/demo cues
• Repository and overview analysis panels now render richer sections instead of a short templated summary, making the output easier to review and adapt for resumes or portfolios
• Export PDF generation now reuses already available Markdown artifacts when possible and prefers richer resume bullets over generic highlights for selected projects
• Benchmark suggestions and GitScore explanations are now more actionable in both the API shape and frontend presentation
• Repository analysis cache versioning was bumped to refresh older, shallow analysis snapshots

[v1.3.1] - 2026-04-17

Added

• UI guidelines regression coverage for page-level accessibility and HTML correctness checks
• Preservation tests covering compare input submission, repository card click separation, overview history navigation, and loading-copy stability

Changed

• Replaced loading and async copy ASCII ellipses with Unicode ellipses across Compare, CompareRepos, Repositories, Export, AIAnalysis, and Overview
• Added missing type, width / height, aria-label, aria-live, role="alert", and chart accessibility attributes across affected frontend pages
• Restructured repository cards to avoid nested interactive elements and converted overview history row activation to semantic buttons
• Updated affected unit and property tests to align with the new UI guidelines and stabilized property-based workspace/app-context test fixtures

[v1.3.0] - 2026-04-15

Added

• OAuth-based GitHub sign-in, token-aware repository fetching, and re-auth prompts for expired sessions
• Server-sent event generation progress, workspace import/export, and repository analysis LRU caching
• Error boundary and offline banner UX safeguards across the frontend
• Property-based regression coverage for request deduplication, XSS sanitization, workspace snapshots, cache bounds, history parsing, GitScore invariants, and social card theming

Changed

• GitScore weighting now includes issue participation, quality density, and higher documentation ceiling
• Export flow now sanitizes rendered resume HTML and delegates PDF generation to the backend POST export endpoint
• Social card rendering now honors light and dark theme selection end to end
• Shared frontend config, i18n wiring, history utilities, and responsive topbar behavior were consolidated for consistency

[v1.2.0] - 2026-04-14

Added

• Frontend workspace isolation with persisted workspaceId, automatic X-Codefolio-Workspace request header, and startup registration via POST /api/workspaces/ensure
• Workspace-aware benchmark persistence so the latest repository comparison result can be reused on the export page without rebuilding query params
• Docker support for the full stack with backend/Dockerfile, frontend/Dockerfile, frontend/nginx.conf, .dockerignore, and root docker-compose.yml

Changed

• Backend snapshot storage now supports tenant_scope and records active workspaces in a dedicated workspaces table
• AI analysis, repository benchmark analysis, Redis cache keys, and snapshot invalidation now respect workspace scope instead of sharing one global artifact set
• Repository comparison flow now fills empty benchmark slots more naturally and stores successful results into the current workspace context

[v1.1.0] - 2026-04-14

新增功能

后端

• benchmark_recommendation_service.py：基于 GitHub Search API 的标杆仓库推荐引擎，按语言、话题、规模分类筛选并排序
• rate_limiter.py：滑动窗口 per-IP 速率限制器，防止 benchmark 接口被滥用
• token_redaction.py：GitHub Token 脱敏工具，确保 token 不出现在日志或错误响应中
• i18n/en.json / i18n/zh.json：benchmark 维度标签、行动项、假设卡片的中英文翻译文件
• database.py：SQLite 数据库辅助模块
• GET /api/repos/suggest-benchmarks：推荐相似标杆仓库接口
• DELETE /api/repos/cache/{owner}/{repo}：手动清除仓库缓存接口

前端

• BenchmarkMatrix.tsx：响应式特征对比表格，支持键盘导航、颜色编码（missing/weak/medium/strong）、行展开查看原始信号
• ActionList.tsx：可排序（优先级/投入/影响）、可筛选（维度）的行动项列表，支持快速收益高亮、时间线分组、localStorage 持久化
• HypothesisCards.tsx：可折叠的成功假设卡片，支持按可迁移性和类别筛选
• benchmarkExport.ts：将 BenchmarkReport 导出为 Markdown（中英文），含特征矩阵表格、行动项编号列表、假设章节
• formatCacheAge.ts：缓存时间格式化与 7 天过期检测

测试覆盖

后端 — 基于属性的测试（Hypothesis，每项 100+ 次迭代）

测试文件	覆盖属性
test_repository_profile_service.py	P2 名称规范化、P3 缓存往返一致性
test_dimension_analyzer.py	P5 维度完整性、P6 单元格结构、边界用例
test_bucket_service.py	P16 分桶逻辑、P10 桶描述存在性
test_action_generator.py	P8 行动项结构与排序、P9 快速收益识别
test_benchmark_infrastructure.py	P7 无 LLM 分析、P11 多仓库矩阵结构、P13 部分失败处理
test_narrative_generation.py	P14 假设证据要求、P15 置信度赋值、P18 单次 LLM 调用、P19 README 截断、P30 LLM 调用次数上限
test_api_endpoints.py	P22 错误响应格式、P12 benchmark 数量上限、集成测试
test_language_handling.py	P28 语言参数处理
test_caching_properties.py	P25 缓存键格式、P26 差异化 TTL、P27 缓存失效、P29 缓存性能、P31 跨请求复用
test_security_properties.py	P34 公开仓库访问、P35 Token 安全、P36 速率限制
test_recommendation_service.py	P20 推荐数量边界、P21 推荐相似性过滤

前端 — 单元测试（Vitest + @testing-library/react）

测试文件	覆盖内容
CompareRepos.test.tsx	输入校验、添加/删除 benchmark、加载/错误状态
BenchmarkMatrix.test.tsx	渲染、颜色编码、键盘导航
ActionList.test.tsx	排序、筛选、快速收益高亮、分组折叠
HypothesisCards.test.tsx	默认折叠、展开/折叠、可迁移性筛选
SuggestionUI.test.tsx	推荐按钮、面板展示、添加推荐、空/错误状态
Export.test.tsx	Markdown 生成（含/不含 benchmark 章节）、导出页面渲染

前端 — 属性测试（fast-check，100+ 次迭代）

测试文件	覆盖属性
urlParams.test.ts	P23 URL 参数隔离（mine/b 与 users 互不干扰）
staleness.test.ts	P33 过期警告显示（7 天阈值）
benchmarkExport.test.ts	P24 Markdown 导出完整性

Bug 修复

• dimension_analyzer.py：修复 analyze_compliance 将 None 的 license_type 直接写入 raw_features，违反 Req 4.10（未检测到的特征应标记为 "not detected" 而非 None）
• benchmark_analysis_service.py：补全缺失的 _build_bucket、_invert_scores 辅助方法

---

[v1.0.0] - 2026-04-13

• 初始发布：仓库 benchmark API、分析服务、CompareRepos UI

v1.2.0

Changelog

[v1.2.0] - 2026-04-14

Added

• Frontend workspace isolation with persisted workspaceId, automatic X-Codefolio-Workspace request header, and startup registration via POST /api/workspaces/ensure
• Workspace-aware benchmark persistence so the latest repository comparison result can be reused on the export page without rebuilding query params
• Docker support for the full stack with backend/Dockerfile, frontend/Dockerfile, frontend/nginx.conf, .dockerignore, and root docker-compose.yml

Changed

• Backend snapshot storage now supports tenant_scope and records active workspaces in a dedicated workspaces table
• AI analysis, repository benchmark analysis, Redis cache keys, and snapshot invalidation now respect workspace scope instead of sharing one global artifact set
• Repository comparison flow now fills empty benchmark slots more naturally and stores successful results into the current workspace context

[v1.1.0] - 2026-04-14

新增功能

后端

• benchmark_recommendation_service.py：基于 GitHub Search API 的标杆仓库推荐引擎，按语言、话题、规模分类筛选并排序
• rate_limiter.py：滑动窗口 per-IP 速率限制器，防止 benchmark 接口被滥用
• token_redaction.py：GitHub Token 脱敏工具，确保 token 不出现在日志或错误响应中
• i18n/en.json / i18n/zh.json：benchmark 维度标签、行动项、假设卡片的中英文翻译文件
• database.py：SQLite 数据库辅助模块
• GET /api/repos/suggest-benchmarks：推荐相似标杆仓库接口
• DELETE /api/repos/cache/{owner}/{repo}：手动清除仓库缓存接口

前端

• BenchmarkMatrix.tsx：响应式特征对比表格，支持键盘导航、颜色编码（missing/weak/medium/strong）、行展开查看原始信号
• ActionList.tsx：可排序（优先级/投入/影响）、可筛选（维度）的行动项列表，支持快速收益高亮、时间线分组、localStorage 持久化
• HypothesisCards.tsx：可折叠的成功假设卡片，支持按可迁移性和类别筛选
• benchmarkExport.ts：将 BenchmarkReport 导出为 Markdown（中英文），含特征矩阵表格、行动项编号列表、假设章节
• formatCacheAge.ts：缓存时间格式化与 7 天过期检测

测试覆盖

后端 — 基于属性的测试（Hypothesis，每项 100+ 次迭代）

测试文件	覆盖属性
test_repository_profile_service.py	P2 名称规范化、P3 缓存往返一致性
test_dimension_analyzer.py	P5 维度完整性、P6 单元格结构、边界用例
test_bucket_service.py	P16 分桶逻辑、P10 桶描述存在性
test_action_generator.py	P8 行动项结构与排序、P9 快速收益识别
test_benchmark_infrastructure.py	P7 无 LLM 分析、P11 多仓库矩阵结构、P13 部分失败处理
test_narrative_generation.py	P14 假设证据要求、P15 置信度赋值、P18 单次 LLM 调用、P19 README 截断、P30 LLM 调用次数上限
test_api_endpoints.py	P22 错误响应格式、P12 benchmark 数量上限、集成测试
test_language_handling.py	P28 语言参数处理
test_caching_properties.py	P25 缓存键格式、P26 差异化 TTL、P27 缓存失效、P29 缓存性能、P31 跨请求复用
test_security_properties.py	P34 公开仓库访问、P35 Token 安全、P36 速率限制
test_recommendation_service.py	P20 推荐数量边界、P21 推荐相似性过滤

前端 — 单元测试（Vitest + @testing-library/react）

测试文件	覆盖内容
CompareRepos.test.tsx	输入校验、添加/删除 benchmark、加载/错误状态
BenchmarkMatrix.test.tsx	渲染、颜色编码、键盘导航
ActionList.test.tsx	排序、筛选、快速收益高亮、分组折叠
HypothesisCards.test.tsx	默认折叠、展开/折叠、可迁移性筛选
SuggestionUI.test.tsx	推荐按钮、面板展示、添加推荐、空/错误状态
Export.test.tsx	Markdown 生成（含/不含 benchmark 章节）、导出页面渲染

前端 — 属性测试（fast-check，100+ 次迭代）

测试文件	覆盖属性
urlParams.test.ts	P23 URL 参数隔离（mine/b 与 users 互不干扰）
staleness.test.ts	P33 过期警告显示（7 天阈值）
benchmarkExport.test.ts	P24 Markdown 导出完整性

Bug 修复

• dimension_analyzer.py：修复 analyze_compliance 将 None 的 license_type 直接写入 raw_features，违反 Req 4.10（未检测到的特征应标记为 "not detected" 而非 None）
• benchmark_analysis_service.py：补全缺失的 _build_bucket、_invert_scores 辅助方法

---

[v1.0.0] - 2026-04-13

• 初始发布：仓库 benchmark API、分析服务、CompareRepos UI

Codefolio v1.1.0

Highlights

• Repository Benchmark fully shipped: 8-dimension gap matrix, action items, success hypothesis cards, optional LLM narrative summary
• Suggest Benchmarks: auto-find similar repos by language, topics, and size category via GitHub Search API
• Markdown export of full benchmark reports (EN / ZH)
• Staleness warning when cached data is older than 7 days
• Rate limiting per IP on benchmark endpoints
• Token redaction GitHub tokens never appear in logs or error responses
• Full test suite: 36 property-based tests (Hypothesis, 100+ iterations each) + frontend unit/property tests (Vitest + fast-check)

What changed

Backend New Services

• benchmark_recommendation_service.py: GitHub Search API based suggestion engine
• rate_limiter.py: sliding-window per-IP rate limiter
• token_redaction.py: GitHub token redaction utility
• i18n/en.json / i18n/zh.json: EN/ZH translation files for benchmark text
• New endpoints: GET /api/repos/suggest-benchmarks, DELETE /api/repos/cache/{owner}/{repo}

Backend Updated Services

• benchmark_analysis_service.py: add _build_hypotheses, _generate_narrative, _build_bucket, partial failure handling
• dimension_analyzer.py: fix None license_type leaking into raw_features (Req 4.10)
• repository_profile_service.py: TTL constants, normalize_repo_name, serialize/deserialize helpers
• cache_keys.py: repo_profile_cache_key, benchmark_result_cache_key, suggestion_cache_key

Frontend New Components

• BenchmarkMatrix.tsx: responsive feature comparison table with keyboard navigation
• ActionList.tsx: sortable/filterable action items with quick-win highlighting
• HypothesisCards.tsx: collapsible hypothesis cards with transferability filter
• benchmarkExport.ts: Markdown export utility (EN/ZH)
• formatCacheAge.ts: cache age formatting and 7-day staleness detection

Frontend Updated Pages

• CompareRepos.tsx: suggest-benchmarks UI, staleness warning, export buttons
• Export.tsx: optional benchmark report section

Tests

• 36 backend property tests (Properties 236) via Hypothesis
• Frontend unit tests: CompareRepos, BenchmarkMatrix, ActionList, HypothesisCards, SuggestionUI, Export
• Frontend property tests: URL isolation (P23), staleness (P33), Markdown export (P24) via fast-check

Run locally

• Backend: Python 3.11+, Redis; see README.md / README.zh-CN.md for .env and startup.
• Frontend: Node 18+; npm install and npm run dev.

License

PolyForm Noncommercial License 1.0.0 see LICENSE.

Codefolio v1.0.0

Highlights

• Repository benchmark: FastAPI router and services for peer-style repo analysis (dimensions, buckets, actions).
• Frontend: CompareRepos page, compare styling updates, benchmark export helper.
• Integration: GitHub service and cache key updates to support the benchmark flow.

Run locally

• Backend: Python 3.11+, Redis; see README.md / README.zh-CN.md for .env and startup.
• Frontend: Node 18+; npm install and npm run dev.

License

PolyForm Noncommercial License 1.0.0 鈥?see LICENSE.