Commit Graph

63 Commits

Author SHA1 Message Date
Albert 1d34e32788 feat: wire PMFlow changeset commit frontend 2026-04-26 09:19:35 +08:00
Albert e04ea4d1ee chore: checkpoint current PMFlow workspace 2026-04-26 09:07:16 +08:00
Yuuuu 2eaf8e9deb Unify structured runtime semantics and internal domain clarify 2026-04-20 20:03:10 +08:00
Yuuuu afa51fc45a feat: unify project operating context and polish Feishu QA handoff 2026-04-18 14:13:34 +08:00
Yuuuu 3f94cf702c Add explicit QA anchors and route governance pages 2026-04-17 14:01:53 +08:00
Yuuuu 4720c8117d feat: add external query and semantic facts runtime 2026-04-16 11:17:52 +08:00
Yuuuu e42652b619 chore: reset repository history 2026-04-14 10:54:06 +08:00
Yuuuu 7f9571c3fa chore: snapshot current code before refactor 2026-03-10 15:18:23 +08:00
Yuuuu 16f30bc2c6 feat: complete PMFlow implementation 2026-02-27 08:56:49 +08:00
Yuuuu 319a422362 feat: 添加 LLM 服务按需启动功能
新增功能:
- 后端新增 /ai/start API,支持启动本地 Ollama
- 后端改进 /ai/status API,返回提供商类型
- 前端看板页面添加启动/停止控制按钮
- 提供商感知 UI:云端 API 无需启动,本地 Ollama 可按需控制

用户体验:
- 云端模式:显示"云端服务已就绪",直接可分析
- 本地模式:显示"启动 Ollama"按钮,按需启动节省内存
- 分析完成后:仅本地模式自动停止 Ollama

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-21 00:01:43 +08:00
Yuuuu 3376c224e3 refactor: 重构设置页面为 Linear.app 风格
设计改进:
- 左侧窄边栏 (200px) + 小图标 (18px)
- 主内容区单列居中 (最大宽度 720px)
- Segmented Control 风格的供应商选择
- 统一表单控件规格 (36px 高度, 6px 圆角)
- 深靛蓝 (#4f46e5) 主色调
- 极简专业的视觉风格

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-20 23:41:30 +08:00
Yuuuu 689cf94d4b feat: 前端添加 LLM 配置设置页面
新增功能:
- 前端设置页面可切换 LLM 提供商
- 支持 DeepSeek, OpenAI, Qwen, Ollama
- API 密钥配置界面
- 配置保存到 .env 文件

后端 API:
- GET /ai/config - 获取当前配置
- POST /ai/config - 更新配置

使用方式:
1. 打开前端 "⚙️ 设置" 页面
2. 选择 LLM 提供商
3. 输入 API 密钥(云端)或设置地址(本地)
4. 保存配置并重启后端

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-20 22:48:38 +08:00
Yuuuu 0981acb06c Initial commit: Neural Sync - AI-native project management platform
Features:
- Signal monitoring with AI strategic analysis
- Experiment management with score tracking
- Context knowledge nodes with RAG
- Multi-model voting for high-impact signals
- Ollama integration for local LLM

Tech stack:
- Backend: FastAPI + SQLAlchemy + PostgreSQL
- Frontend: Vue 3 + Naive UI + TypeScript
- AI: Ollama (llama3, nomic-embed-text)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-20 18:06:39 +08:00