For most teams, WeCom development is not about building a new system from scratch. They face a more common and more realistic scenario: the business system already exists and has been running for years — the attendance module went live long ago, and the approval workflow is built on Activiti with complex processes such as countersign (all must approve), or-sign (any one approves), and organization-chart-based multi-level approval, and the approval process queries and interacts with attendance data along the way. The requirement now is: bring this system into WeCom, so employees can open it from the WeCom Workbench and use it immediately — no username or password needed; they land straight on their own attendance data and to-do tasks.
Under these conditions, the H5 app model is often a better fit than a Mini Program: the existing system is already an Angular + SpringBoot web application, so H5 can reuse the frontend pages and backend APIs directly, and with WeCom OAuth2 web authorization (snsapi_base) it achieves a fully silent automatic login (SSO). Deployment takes effect immediately with no review or release, and iteration cost is minimal when approval forms change frequently.
Set against the backdrop of “an existing attendance management system + a complex Activiti approval workflow,” and taking the H5 model as the main thread, this article explains systematically how to complete the WeCom-side integration without rewriting the business system. It focuses on the full chain of OAuth2 silent automatic login, the binding and mapping between WeCom accounts and system accounts, JS-SDK device capability invocation, and the WeCom-side implementation of Activiti countersign / or-sign / organization-chart approval interacting with attendance (to-do push, one-tap approval cards, organization-chart synchronization).
OpenClaw is an open-source personal AI assistant framework that runs on your own devices. Built on TypeScript, it connects the messaging channels you use every day through a Gateway, letting you chat with your AI assistant on 20+ platforms including WhatsApp, Telegram, Slack, and Discord.
The goal of this article is the same as the Hermes guide: read just this one article, and you’ll master OpenClaw. From installation and deployment to usage tips, it’s all covered.
Hermes Agent is an open-source AI Agent framework developed by Nous Research that runs in terminals, messaging platforms, and IDEs. It can interact with your system through tool calls, supports 20+ LLM providers, and runs on Linux, macOS, Windows, and Docker.
The goal of this article is simple: read just this one piece, and you’ll be able to master Hermes. From installation and deployment to usage tips, it’s all covered.
After developing a WeChat Mini Program, how to deploy it is the first challenge many developers face. This article provides a detailed walkthrough of deploying a typical WeChat Mini Program project...
PostgreSQL’s Foreign Table is a powerful feature that allows us to query remote data sources as if they were local tables. However, in practice, especially when involving LEFT JOIN with Foreign Table primary keys, performance issues often become a headache. This article will deeply analyze the principles of Foreign Table and provide systematic performance optimization solutions.
With the advancement of AI technology, more development work can be assisted by AI. However, a single AI assistant often struggles to handle all roles in complex projects — product managers need to understand requirements, architects need to design systems, developers need to write code, and QA needs to verify quality.
Hermes’s Profile mechanism allows us to define different identities and responsibilities for AI, building a complete AI development team. This article details how to set up and use this system.
Spring Security provides three types of annotations for method-level authorization control:
- JSR-250 Annotations: Standard Java EE specification annotations
-