Compare commits

..
9 Commits
6 changed files with 563 additions and 124 deletions
+266
View File
@@ -0,0 +1,266 @@
<picture src="https://img.shields.io/badge/python-3.11%2B-blue" alt="Python" /><picture src="https://img.shields.io/badge/vue-3.4%2B-4FC08D" alt="Vue" /><picture src="https://img.shields.io/badge/license-MIT-green" alt="License" />
# TaskPulse — AI Agent Scheduled Task Monitor
> **Let your AI agents self-report. You just glance at the dashboard.**
TaskPulse is a tracking and monitoring platform designed for AI agents' scheduled tasks. Unlike traditional CRUD-heavy admin panels, TaskPulse is **AI-native**: give your agents an API spec, and they autonomously register themselves, report execution results, and update their status — no manual form-filling required.
[📖 中文文档](./README.md)
---
## Philosophy
```
┌─────────────────────┐
│ TaskPulse Dashboard│
│ Unified Status View│
└──────┬──────────┬───┘
│ │
┌─────────────────┘ └─────────────────┐
▼ ▼
┌─────────────────┐ ┌──────────────────────┐
│ AI Agent A │ register tasks → execute │ AI Agent B │
│ ├─ data sync │ ← report result + log │ ├─ order fetch │
│ └─ report gen │ │ └─ inventory sync │
└─────────────────┘ └──────────────────────┘
```
**AI-Native by Design**
Traditional schedulers require humans to fill forms and manage cron expressions. But AI agents can read API docs and act autonomously:
1. **Self-Register** — One API call to register both the agent and all its tasks
2. **Report Results** — After each execution, POST back the result with logs and duration
3. **Auto-Alert** — If a task misses its window, the system alerts automatically
The only thing humans need to do: **open the dashboard and see the big picture.**
---
## Features
### 🤖 AI-Native Onboarding
- **Batch Registration**: `POST /api/agents/register-with-tasks` — agent + tasks in one shot
- **API Key Auth**: Each agent gets its own key for reporting
- **Re-registration**: Deleted agents can re-register anytime
- **Agent Avatars**: First-letter + deterministic color avatars for quick visual identification
### 📊 Unified Dashboard
- **Summary Cards**: agent count, task count, 24h executions, pending alerts
- **Global Task View**: all tasks from all agents in one place
- **Task Tags**: label tasks (e.g. "data-sync", "critical"), filter by tags
- **Human-Readable Cron**: `*/5 * * * *` → "Every 5 minutes"
- **Localized Time**: all timestamps auto-converted from UTC to browser timezone
### ⏱ Execution Tracking
- **Live Reporting**: agents POST results immediately after execution
- **Execution History**: full timeline per task, paginated
- **Log Viewer**: dark terminal-style viewer for debugging
- **Search & Filter**: by task name, tags, and status
### 🚨 Timeout Alerts & Notifications
- **Auto-Scan**: background scheduler checks overdue tasks every 60s
- **Grace Window**: per-task tolerance to avoid false alerts
- **Multi-Channel**: Feishu (Lark) webhook, email SMTP, generic webhook
- **Alert Acknowledgment**: mark alerts as handled for closed-loop management
### 🎨 Modern UI
- **Dark Theme**: gradient backgrounds + neon accents
- **Full-Width Layout**: maximizes screen usage
- **Responsive**: adapts to different screen sizes
---
## Tech Stack
| Layer | Technology |
|---|---|
| Backend | Python FastAPI (async) |
| Production Server | Gunicorn + Uvicorn Worker |
| Frontend | Vue 3 + Element Plus |
| Build Tool | Vite |
| Database | MySQL 8.0 |
| Validation | Pydantic v2 |
| ORM | SQLAlchemy 2.0 (async) |
| Auth | JWT (python-jose) + bcrypt |
| Scheduler | APScheduler (async loop) |
| Deployment | Monolithic (FastAPI serves SPA) |
---
## Quick Start
### Local Development
```bash
# 1. Clone
git clone https://github.com/oventh/TaskPulse.git
cd TaskPulse
# 2. Backend
cd backend
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -r requirements.txt
cp .env.example .env # edit .env with your DB config
uvicorn app.main:app --reload --port 8000
# 3. Frontend (new terminal)
cd frontend
npm install
npm run dev # http://localhost:3000
```
> Frontend dev server proxies `/api` to `localhost:8000`.
### Production Deployment
```bash
# 1. Build frontend
cd frontend && npm install && npm run build
# 2. Deploy to server
rsync -avz --exclude='.venv' --exclude='node_modules' --exclude='.git' \
./ user@server:/home/app/taskpulse/
# 3. Install deps + start with Gunicorn
cd /home/app/taskpulse/backend
python3 -m venv .venv && .venv/bin/pip install -r requirements.txt
.venv/bin/gunicorn app.main:app \
-k uvicorn.workers.UvicornWorker \
-b 0.0.0.0:8000 -w 2 --preload --daemon
```
---
## Agent Integration
Let your AI agent self-register in 3 steps:
```python
import requests
BASE = "https://your-domain.com"
# 1. Register agent + all its tasks (one API call)
resp = requests.post(f"{BASE}/api/agents/register-with-tasks", json={
"name": "my-agent",
"description": "My AI Agent",
"tasks": [
{"name": "sync-data", "cron_expression": "*/5 * * * *"},
{"name": "daily-report", "cron_expression": "0 9 * * *"}
]
})
data = resp.json()
API_KEY = data["agent"]["api_key"] # ← save this
AGENT_ID = data["agent"]["id"]
# 2. Report execution result after each run
requests.post(f"{BASE}/api/tasks/{TASK_ID}/executions",
headers={"Authorization": f"Bearer {API_KEY}"},
json={"status": "success", "duration_ms": 1500, "log": "Task completed"})
```
> For curl, batch registration, and more examples, see the Chinese doc [AGENT_GUIDE.md](./AGENT_GUIDE.md).
---
## Configuration
| Variable | Default | Description |
|----------|---------|-------------|
| `TASKPULSE_DB_HOST` | localhost | MySQL host |
| `TASKPULSE_DB_PORT` | 3306 | MySQL port |
| `TASKPULSE_DB_USER` | dbuser | Database user |
| `TASKPULSE_DB_PASSWORD` | - | Database password |
| `TASKPULSE_DB_NAME` | taskpulse | Database name |
| `TASKPULSE_DEBUG` | true | Debug mode (false in production) |
| `TASKPULSE_SECRET_KEY` | - | JWT signing key |
| `TASKPULSE_SMTP_*` | - | Email notification config |
See `backend/.env.example` for the full list.
---
## API Overview
### Agents
| Method | Path | Description |
|--------|------|-------------|
| POST | `/api/agents` | Register an agent (returns API Key) |
| **POST** | **`/api/agents/register-with-tasks`** | **🌟 Batch register agent + tasks** |
| GET | `/api/agents` | List all agents |
| PUT | `/api/agents/{id}` | Update agent name/description |
| DELETE | `/api/agents/{id}` | Delete agent (cascades to tasks) |
### Tasks
| Method | Path | Description |
|--------|------|-------------|
| POST | `/api/tasks?agent_id=X` | Create a scheduled task |
| GET | `/api/tasks` | List tasks (`?q=&tags=&status=` filters) |
| PUT | `/api/tasks/{id}` | Update task name/tags/description |
| DELETE | `/api/tasks/{id}` | Delete a task |
### Execution Reporting
| Method | Path | Description |
|--------|------|-------------|
| POST | `/api/tasks/{id}/executions` | **Agent reports execution result** |
| GET | `/api/tasks/{id}/executions` | View execution history (paginated) |
| GET | `/api/tasks/{id}/executions/{eid}` | View single execution detail |
### Dashboard & Config
| Method | Path | Description |
|--------|------|-------------|
| GET | `/api/dashboard/summary` | Dashboard summary stats |
| GET | `/api/dashboard/tasks` | Dashboard full task view |
| GET | `/api/system/config` | Get system config |
| PUT | `/api/system/config` | Update system config (Base URL) |
### Alerts & Notifications
| Method | Path | Description |
|--------|------|-------------|
| GET | `/api/alerts` | List alerts |
| POST | `/api/alerts/{id}/acknowledge` | Acknowledge an alert |
| POST | `/api/notification-channels` | Add notification channel |
Full OpenAPI docs at `/docs` (Swagger UI).
---
## Project Structure
```
taskpulse/
├── backend/
│ ├── app/
│ │ ├── main.py # FastAPI entry + SPA static files + lifecycle
│ │ ├── config.py # Environment variables (pydantic-settings)
│ │ ├── database.py # Async MySQL connection
│ │ ├── models/ # SQLAlchemy models (6 tables)
│ │ ├── schemas/ # Pydantic request/response models
│ │ ├── api/ # REST API routes (7 modules)
│ │ └── services/ # Business logic layer
│ │ ├── scheduler.py # ⏰ Background scheduler (timeout detection)
│ │ └── notification.py # 📢 Feishu/Email/Webhook sender
│ ├── alembic/ # DB migrations
│ └── requirements.txt
├── frontend/
│ ├── src/
│ │ ├── views/ # 6 pages (dark theme)
│ │ ├── router/ # Route guards + login redirect
│ │ ├── api/ # Axios client + 401 interceptor
│ │ └── assets/ # Global styles + Element Plus overrides
│ └── package.json
├── AGENT_GUIDE.md # Agent API guide (Chinese)
├── README.md # Documentation (Chinese)
└── README.en.md # Documentation (English)
```
---
## License
MIT License © 2026
+189 -110
View File
@@ -1,28 +1,96 @@
# TaskPulse — AI Agent Task Monitor
<picture src="https://img.shields.io/badge/python-3.11%2B-blue" alt="Python" /><picture src="https://img.shields.io/badge/vue-3.4%2B-4FC08D" alt="Vue" /><picture src="https://img.shields.io/badge/license-MIT-green" alt="License" />
统一看板,监控 AI Agent 定时任务的运行状态、执行历史,并在任务超时未执行时发送告警通知。
# TaskPulse — AI Agent 定时任务监控系统
**线上地址**: https://task.pags.cn
> **让 AI Agent 自己汇报工作,你只需要打开看板看一眼。**
TaskPulse 是一个面向 AI Agent 的定时任务跟踪与监控平台。它不是传统的"你去填表创建任务"的后台管理系统,而是**给 AI Agent 一份 API 说明书,让 Agent 自主注册、自动汇报**——你只需要在统一的 Dashboard 上看全局状态。
## 系统截图
<img src="https://pub-44c5bd2a850e4bc7aab6f5f8701493fd.r2.dev/images/111.png"></img>
[🌐 English](./README.en.md)
---
## 设计理念
```
┌─────────────────────┐
│ TaskPulse Dashboard│
│ 统一状态看板 │
└──────┬──────────┬───┘
│ │
┌─────────────────┘ └─────────────────┐
▼ ▼
┌─────────────────┐ ┌──────────────────────┐
│ AI Agent A │ 注册任务 → 定时执行 │ AI Agent B │
│ └─ 数据同步 │ ← 汇报结果 + 日志 │ └─ 订单采集 │
│ └─ 报表生成 │ │ └─ 库存同步 │
└─────────────────┘ └──────────────────────┘
```
**核心理念:AI 原生接入**
传统定时任务管理需要用户手动填写 cron 表达式、配置参数。但 AI Agent 本身就具备理解和执行能力——只需给 Agent 一份 API 文档,它就能:
1. **自主注册** — 通过一条 API 把自己和所有定时任务登记到系统
2. **按时汇报** — 每次任务执行完毕,主动 POST 结果(含日志、耗时、状态)
3. **自动纠错** — 如果某个任务到点没执行,系统自动告警
人类要做的只有一件事:**打开看板,看一眼全局状态。**
---
## 核心功能
- **AI 原生接入** — AI Agent 通过 API 一次性注册自己 + 所有定时任务,无需手动填写表单
- **统一状态看板** — 所有定时任务的运行状态、最近运行时间、结果一目了然
- **执行汇报** — Agent 执行完任务后主动汇报结果,含详细日志
- **超时告警** — 任务未按时执行时自动告警
- **多渠道通知** — 支持飞书 Webhook、邮件、通用 Webhook
- **Base URL 可配置** — 系统设置页面可配置公网访问地址,所有接入指令自动更新
### 🤖 AI 原生接入
- **一次性批量注册**`POST /api/agents/register-with-tasks` — Agent 在一分钟内完成注册 + 所有任务登记
- **API Key 鉴权**:每个 Agent 独立密钥,汇报时自动校验
- **无缝重注册**:删除后可重新注册,获取新 Key 继续工作
- **Agent 头像**:名称首字 + 确定性色彩头像,多 Agent 一目了然
### 📊 统一状态看板
- **概览卡片**:Agent 数量、任务总数、24h 执行量、待处理告警
- **全任务视图**:所有 Agent 的任务集中展示,状态、最近运行、结果一目了然
- **任务标签**:可对任务打标签(如"数据同步"、"重要"),按标签筛选
- **Cron 人类可读**`*/5 * * * *` 自动显示为 "每5分钟"
- **时间本地化**:所有时间自动从 UTC 转为浏览器本地时区
### ⏱ 执行汇报与追踪
- **实时汇报**Agent 执行完任务后 POST 结果,系统记录执行日志
- **执行历史**:每个任务都有完整的时间线,支持分页查看
- **日志查看器**:暗色终端风格,方便排查问题
- **搜索过滤**:按任务名称、标签、状态快速筛选
### 🚨 超时告警与通知
- **自动检测**:后台调度器每 60 秒扫描未按时执行的任务
- **宽容窗口**:每个任务可单独配置容忍时间,避免误报
- **多渠道通知**:飞书 Webhook、邮件 SMTP、通用 Webhook
- **告警确认**:已处理的告警可标记确认,闭环管理
### 🎨 现代化界面
- **暗色主题**:深色渐变背景 + 霓虹色点缀,降低视觉疲劳
- **全宽布局**:充分利用屏幕空间
- **响应式**:适配不同分辨率
---
## 技术栈
| 层 | 技术 |
|---|---|
| 后端 | Python FastAPI (异步) + Gunicorn (生产) |
| 前端 | Vue3 + Element Plus (暗色主题) |
| 后端框架 | Python FastAPI (异步) |
| 生产服务 | Gunicorn + Uvicorn Worker |
| 前端框架 | Vue 3 + Element Plus |
| 构建工具 | Vite |
| 数据库 | MySQL 8.0 |
| 部署 | 单体应用(FastAPI serve Vue3 构建产物) |
| 数据校验 | Pydantic v2 |
| ORM | SQLAlchemy 2.0 (异步) |
| 认证 | JWT (python-jose) + bcrypt |
| 定时检测 | APScheduler (异步) |
| 部署形态 | 单体应用 (FastAPI serve 前端静态文件) |
---
## 快速开始
@@ -30,88 +98,55 @@
```bash
# 1. 克隆项目
git clone https://gitea.tatta.cn/Tatta/TaskPulse.git
git clone https://github.com/oventh/TaskPulse.git
cd TaskPulse
# 2. 后端 - 创建虚拟环境并安装依赖
# 2. 后端
cd backend
python -m venv .venv
# Windows: .venv\Scripts\activate
# Linux: source .venv/bin/activate
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -r requirements.txt
# 3. 复制环境变量并修改数据库配置
cp .env.example .env
# 编辑 .env,配置 MySQL 连接信息
# 4. 启动后端开发服务器
cp .env.example .env # 编辑 .env 配置数据库
uvicorn app.main:app --reload --port 8000
# 5. 前端(新终端)
# 3. 前端(新终端)
cd frontend
npm install
npm run dev # 访问 http://localhost:3000
npm run dev # http://localhost:3000
```
> 前端 dev server 自动代理 `/api` 请求到 `localhost:8000`
> 前端 `npm run dev` 会代理 `/api` 到 `localhost:8000`
### 生产部署
```bash
# 1. 构建前端
cd frontend
npm install && npm run build
cd frontend && npm install && npm run build
# 2. 部署到服务器
# 将整个项目复制到服务器(排除 .venv 和 node_modules
rsync -avz --exclude='.venv' --exclude='node_modules' --exclude='.git' \
./ user@server:/home/openclaw/app/taskpulse/
./ user@server:/home/app/taskpulse/
# 3. 在服务器上创建虚拟环境并安装依赖
cd /home/openclaw/app/taskpulse/backend
python3 -m venv .venv
.venv/bin/pip install -r requirements.txt
# 4. 使用 Gunicorn 启动
# 3. 安装依赖 + Gunicorn 启动
cd /home/app/taskpulse/backend
python3 -m venv .venv && .venv/bin/pip install -r requirements.txt
.venv/bin/gunicorn app.main:app \
-k uvicorn.workers.UvicornWorker \
-b 0.0.0.0:8000 \
-w 2 \
--access-logfile /var/log/taskpulse/access.log \
--error-logfile /var/log/taskpulse/error.log \
--daemon
-b 0.0.0.0:8000 -w 2 --preload --daemon
```
## 环境变量
---
复制 `backend/.env.example``backend/.env` 并按需修改:
## Agent 接入指南
| 变量 | 默认值 | 说明 |
|------|--------|------|
| TASKPULSE_DB_HOST | localhost | MySQL 主机 |
| TASKPULSE_DB_PORT | 3306 | MySQL 端口 |
| TASKPULSE_DB_USER | dbuser | 数据库用户 |
| TASKPULSE_DB_PASSWORD | - | 数据库密码 |
| TASKPULSE_DB_NAME | taskpulse | 数据库名 |
| TASKPULSE_DEBUG | true | 调试模式(生产环境设为 false) |
| TASKPULSE_FEISHU_WEBHOOK_URL | (空) | 飞书 Webhook 地址 |
| TASKPULSE_SMTP_HOST | (空) | SMTP 服务器 |
| TASKPULSE_SMTP_PORT | 587 | SMTP 端口 |
| TASKPULSE_SMTP_USER | (空) | SMTP 用户 |
| TASKPULSE_SMTP_PASSWORD | (空) | SMTP 密码 |
## Agent 接入
详细接入说明见 [AGENT_GUIDE.md](./AGENT_GUIDE.md)。
### 最小接入示例
让 AI Agent 自动接入只需 3 步:
```python
import requests
BASE = "https://task.pags.cn" # 替换为你的实际地址
BASE = "https://your-domain.com" # 替换为你的实际地址
# 1. AI Agent 注册自己 + 所有定时任务(一次性)
# 1. Agent 一次性注册自己 + 所有定时任务
resp = requests.post(f"{BASE}/api/agents/register-with-tasks", json={
"name": "my-agent",
"description": "我的 AI Agent",
@@ -120,38 +155,81 @@ resp = requests.post(f"{BASE}/api/agents/register-with-tasks", json={
{"name": "daily-report", "cron_expression": "0 9 * * *"}
]
})
agent = resp.json()
API_KEY = agent["api_key"] # 保存此 Key
AGENT_ID = agent["agent"]["id"]
data = resp.json()
API_KEY = data["agent"]["api_key"] # 保存此 Key
AGENT_ID = data["agent"]["id"]
# 2. 每次执行汇报结果
# 2. 每次执行完任务后,汇报结果
requests.post(f"{BASE}/api/tasks/{TASK_ID}/executions",
headers={"Authorization": f"Bearer {API_KEY}"},
json={"status": "success", "duration_ms": 1200, "log": "done"})
json={"status": "success", "duration_ms": 1500, "log": "执行完毕"})
```
> 完整接入说明见 [AGENT_GUIDE.md](./AGENT_GUIDE.md),含 curl / Python / 批量注册示例。
---
## 系统配置
| 变量 | 默认值 | 说明 |
|------|--------|------|
| `TASKPULSE_DB_HOST` | localhost | MySQL 主机 |
| `TASKPULSE_DB_PORT` | 3306 | MySQL 端口 |
| `TASKPULSE_DB_USER` | dbuser | 数据库用户 |
| `TASKPULSE_DB_PASSWORD` | - | 数据库密码 |
| `TASKPULSE_DB_NAME` | taskpulse | 数据库名 |
| `TASKPULSE_DEBUG` | true | 调试模式(生产 false |
| `TASKPULSE_SECRET_KEY` | - | JWT 签名密钥 |
| `TASKPULSE_SMTP_*` | - | 邮件通知配置 |
完整变量列表见 `backend/.env.example`
---
## API 概览
### Agent 管理
| 方法 | 路径 | 说明 |
|------|------|------|
| POST | /api/agents | 注册 Agent |
| POST | /api/agents/register-with-tasks | **AI Agent 一次性注册自己 + 所有任务** |
| GET | /api/agents | 列出所有 Agent |
| POST | /api/tasks?agent_id=X | 注册定时任务 |
| GET | /api/tasks | 列出所有任务 |
| DELETE | /api/tasks/{id} | 删除任务 |
| POST | /api/tasks/{id}/executions | 汇报执行结果 |
| GET | /api/tasks/{id}/executions | 查看执行历史 |
| GET | /api/dashboard/summary | 看板概览 |
| GET | /api/dashboard/tasks | 看板任务视图 |
| GET | /api/system/config | 读取系统配置(Base URL |
| PUT | /api/system/config | 更新系统配置 |
| GET | /api/alerts | 告警列表 |
| POST | /api/alerts/{id}/acknowledge | 确认告警 |
| POST | /api/notification-channels | 添加通知渠道 |
| GET | /api/notification-channels | 列出通知渠道 |
| POST | `/api/agents` | 注册 Agent(返回 API Key |
| **POST** | **`/api/agents/register-with-tasks`** | **🌟 Agent + 任务一次性批量注册** |
| GET | `/api/agents` | 列出所有 Agent |
| PUT | `/api/agents/{id}` | 修改 Agent 名称/描述 |
| DELETE | `/api/agents/{id}` | 删除 Agent(关联任务一并删除) |
完整 API 文档见 `/docs`Swagger UI)。
### 定时任务
| 方法 | 路径 | 说明 |
|------|------|------|
| POST | `/api/tasks?agent_id=X` | 注册定时任务 |
| GET | `/api/tasks` | 任务列表(支持 `?q=&tags=&status=` 搜索) |
| PUT | `/api/tasks/{id}` | 修改任务名称/标签/描述 |
| DELETE | `/api/tasks/{id}` | 删除任务 |
### 执行汇报
| 方法 | 路径 | 说明 |
|------|------|------|
| POST | `/api/tasks/{id}/executions` | Agent 汇报执行结果 |
| GET | `/api/tasks/{id}/executions` | 查看执行历史(分页) |
| GET | `/api/tasks/{id}/executions/{eid}` | 查看单次执行详情 |
### 看板与配置
| 方法 | 路径 | 说明 |
|------|------|------|
| GET | `/api/dashboard/summary` | 看板概览统计数据 |
| GET | `/api/dashboard/tasks` | 看板全任务视图 |
| GET | `/api/system/config` | 获取系统配置 |
| PUT | `/api/system/config` | 更新系统配置(Base URL |
### 告警与通知
| 方法 | 路径 | 说明 |
|------|------|------|
| GET | `/api/alerts` | 告警列表 |
| POST | `/api/alerts/{id}/acknowledge` | 确认告警 |
| POST | `/api/notification-channels` | 添加通知渠道(飞书/邮件/Webhook) |
完整 OpenAPI 文档在 `/docs`Swagger UI)查看。
---
## 项目结构
@@ -159,35 +237,36 @@ requests.post(f"{BASE}/api/tasks/{TASK_ID}/executions",
taskpulse/
├── backend/
│ ├── app/
│ │ ├── main.py # FastAPI 入口 + SPA 静态文件服务
│ │ ├── config.py # 环境变量配置
│ │ ├── database.py # 异步数据库连接
│ │ ├── models/ # SQLAlchemy 数据模型
│ │ │ ├── agent.py # Agent(含 API Key 自动生成)
│ │ │ ├── task.py # 定时任务cron、宽容窗口)
│ │ │ ├── execution.py # 执行记录(日志、结果、耗时)
│ │ │ ├── notification.py # 通知渠道告警规则、告警历史
│ │ │ ── system_config.py # 系统配置key-value
│ │ ├── schemas/ # Pydantic 数据校验
│ │ ├── api/ # API 路由
│ │ ├── agents.py # Agent CRUD + 批量注册
│ │ │ ├── tasks.py # 任务 CRUD
│ │ ├── executions.py # 执行汇报 + 历史查询
│ │ ── notifications.py # 通知渠道 + 告警规则 + 告警确认
│ │ │ ├── dashboard.py # 看板概览统计
│ │ │ └── system.py # 系统配置
│ │ └── services/ # 业务逻辑
│ │ ├── agent.py / task.py / execution.py
│ │ ├── scheduler.py # 后台扫描:检测超时任务→生成告警→通知
│ │ └── notification.py # 飞书 / 邮件 / Webhook 通知发送
│ │ ├── main.py # FastAPI 入口 + SPA 静态文件 + 生命周期
│ │ ├── config.py # 环境变量配置 (pydantic-settings)
│ │ ├── database.py # 异步 MySQL 连接
│ │ ├── models/ # SQLAlchemy 数据模型6 张表)
│ │ │ ├── agent.py # Agent + API Key
│ │ │ ├── task.py # 定时任务 + cron + tags
│ │ │ ├── execution.py # 执行记录 + 日志
│ │ │ ├── notification.py # 通知渠道 + 告警规则 + 告警
│ │ │ ── system_config.py # 系统配置 (key-value)
│ │ │ └── user.py # 看板登录用户
│ │ ├── schemas/ # Pydantic 请求/响应模型
│ │ ├── api/ # REST API 路由(7 个模块)
│ │ └── services/ # 业务逻辑层
│ │ ├── scheduler.py # ⏰ 后台调度器(超时检测)
│ │ ── notification.py # 📢 飞书/邮件/Webhook 发送
│ ├── alembic/ # 数据库迁移
│ └── requirements.txt
├── frontend/
│ ├── src/
│ │ ├── views/ # 页面组件(暗色主题)
│ │ ├── router/ # 路由
│ │ ── api/ # Axios API 客户端
│ │ ├── views/ # 6 个页面(暗色主题)
│ │ ├── router/ # 路由守卫 + 登录跳转
│ │ ── api/ # Axios 客户端 + 401 拦截
│ │ └── assets/ # 全局样式 + Element Plus 覆盖
│ └── package.json
├── AGENT_GUIDE.md # Agent 接入指南
├── AGENT_GUIDE.md # Agent API 接入说明
└── README.md
```
---
## 许可证
MIT License © 2026
+26 -5
View File
@@ -1,15 +1,25 @@
"""API router — TaskExecution (reporting + query)."""
from fastapi import APIRouter, Depends, HTTPException, Query
from pydantic import BaseModel
from sqlalchemy import select, func
from sqlalchemy.ext.asyncio import AsyncSession
from app.database import get_db
from app.schemas import ExecutionOut, ExecutionReport
from app.services import ExecutionService, TaskService, AgentService
from app.models import TaskExecution
router = APIRouter(prefix="/api/tasks/{task_id}/executions", tags=["executions"])
class PaginatedExecutions(BaseModel):
items: list[ExecutionOut]
total: int
page: int
page_size: int
@router.post("", response_model=ExecutionOut, status_code=201)
async def report_execution(task_id: int, body: ExecutionReport,
db: AsyncSession = Depends(get_db)):
@@ -46,16 +56,27 @@ async def report_execution(task_id: int, body: ExecutionReport,
return ExecutionOut(**record.__dict__)
@router.get("", response_model=list[ExecutionOut])
@router.get("", response_model=PaginatedExecutions)
async def list_executions(
task_id: int,
limit: int = Query(50, ge=1, le=200),
offset: int = Query(0, ge=0),
page: int = Query(1, ge=1),
page_size: int = Query(20, ge=1, le=100),
db: AsyncSession = Depends(get_db),
):
exec_svc = ExecutionService(db)
records = await exec_svc.get_executions(task_id, limit=limit, offset=offset)
return [ExecutionOut(**r.__dict__) for r in records]
offset = (page - 1) * page_size
records = await exec_svc.get_executions(task_id, limit=page_size, offset=offset)
# Get total count
count_stmt = select(func.count()).select_from(TaskExecution).where(TaskExecution.task_id == task_id)
total = (await db.execute(count_stmt)).scalar() or 0
return PaginatedExecutions(
items=[ExecutionOut(**r.__dict__) for r in records],
total=total,
page=page,
page_size=page_size,
)
@router.get("/recent", response_model=list[ExecutionOut])
+1 -1
View File
@@ -42,7 +42,7 @@
<script setup>
import { ref, onMounted } from 'vue'
import { beijing } from '../utils/time.js'
import dayjs from 'dayjs'
import { listAlerts, acknowledgeAlert } from '../api/index.js'
const alerts = ref([])
+1 -1
View File
@@ -152,7 +152,7 @@
<script setup>
import { ref, onMounted, computed } from 'vue'
import { beijing } from '../utils/time.js'
import dayjs from 'dayjs'
import { getDashboardSummary, getDashboardTasks, getSystemConfig } from '../api/index.js'
const summary = ref({
+78 -5
View File
@@ -72,6 +72,20 @@
</tbody>
</table>
</div>
<!-- Pagination -->
<div class="pagination" v-if="total > pageSize && totalPages > 1">
<span class="page-info"> {{ total }} </span>
<div class="page-btns">
<button class="page-btn" :disabled="currentPage <= 1" @click="goPage(currentPage - 1)">上一页</button>
<template v-for="p in (totalPages || 1)" :key="p">
<button v-if="Math.abs(p - currentPage) <= 2 || p === 1 || p === totalPages"
class="page-btn" :class="{ active: p === currentPage }"
@click="goPage(p)">{{ p }}</button>
<span v-else-if="p === currentPage - 3 || p === currentPage + 3" class="page-ellipsis"></span>
</template>
<button class="page-btn" :disabled="currentPage >= totalPages" @click="goPage(currentPage + 1)">下一页</button>
</div>
</div>
</div>
<!-- Log Dialog -->
@@ -90,9 +104,9 @@
</template>
<script setup>
import { ref, onMounted } from 'vue'
import { ref, onMounted, computed } from 'vue'
import { useRoute } from 'vue-router'
import { beijing } from '../utils/time.js'
import dayjs from 'dayjs'
import { getTask, listExecutions } from '../api/index.js'
const route = useRoute()
@@ -103,14 +117,37 @@ const logVisible = ref(false)
const currentLog = ref('')
const logExecution = ref(null)
// Pagination
const currentPage = ref(1)
const pageSize = ref(20)
const total = ref(0)
const totalPages = computed(() => Math.ceil(total.value / pageSize.value) || 1)
const loadExecutions = async (page) => {
try {
const res = await listExecutions(taskId, { page, page_size: pageSize.value })
const data = res.data || {}
executions.value = data.items || []
total.value = data.total || 0
currentPage.value = data.page || page
} catch (e) {
console.error(e)
executions.value = []
}
}
const goPage = (p) => {
if (p < 1 || p > totalPages.value) return
loadExecutions(p)
}
onMounted(async () => {
try {
const [taskRes, execRes] = await Promise.all([
const [taskRes] = await Promise.all([
getTask(taskId),
listExecutions(taskId, { limit: 100 }),
])
task.value = taskRes.data
executions.value = execRes.data
await loadExecutions(1)
} catch (e) {
console.error(e)
}
@@ -243,4 +280,40 @@ const showLog = (row) => {
word-break: break-all;
margin: 0;
}
/* ── Pagination ──────────────────────────── */
.pagination {
display: flex;
justify-content: space-between;
align-items: center;
padding: 12px 20px;
border-top: 1px solid var(--border-color);
}
.page-info {
font-size: 12px;
color: var(--text-muted);
}
.page-btns {
display: flex;
align-items: center;
gap: 4px;
}
.page-btn {
padding: 4px 10px;
border-radius: 6px;
border: 1px solid var(--border-color);
background: transparent;
color: var(--text-secondary);
font-size: 12px;
cursor: pointer;
transition: all 0.15s;
}
.page-btn:hover { border-color: var(--accent-1); color: var(--accent-1); }
.page-btn.active {
background: var(--accent-1);
border-color: var(--accent-1);
color: #fff;
}
.page-btn:disabled { opacity: 0.3; cursor: not-allowed; }
.page-ellipsis { color: var(--text-muted); font-size: 12px; padding: 0 2px; }
</style>