Skip to content

Version 1.0.0 - "Phoenix"

Choose a tag to compare

View all tags
@bteamapp bteamapp released this 17 Jul 09:09
· 26 commits to main since this release
Phoenix
84e95f0
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Release Notes: Version 1.0.0 - "Phoenix"

Date: 17/07/2025
Codename: Phoenix

This is the first stable version of AI Assistant for LMS hosted on Blogger, an intelligent Q&A system designed to integrate seamlessly into Blogger's TTKT posts. This version focuses on delivering a seamless, secure, and production-ready Q&A experience.

Key Features

  • Integrated AI Q&A based on context: The system uses Google Gemini to analyze the content of a specific article and answer students' questions based only on that context, ensuring accuracy and relevance to the lesson content.

  • Iframe Chat Interface: Provides a modern, responsive chat interface that can be easily embedded into Blogger through a simple JavaScript code. The interface is friendly to both desktop and mobile.

  • Dynamic Data Crawl: The API is capable of automatically crawling and extracting the main content of the article from the provided URL, removing redundant components such as navigation bars, sidebars, and footers.

  • Security based on Whitelist Domain: The core security mechanism is based on authenticating the URL of the article. Only domains pre-configured in allowedDomains are allowed to use the AI service, effectively preventing API abuse from unwanted websites.

  • Health Check Endpoint: Provide the /healthz endpoint for hosting services (such as Render) to monitor the application's health status, ensuring high availability.

Security and Stability Improvements (Improvements & Fixes)

  • FIXED - Not allowed by CORS: Completely fix the CORS bug by changing the security strategy. Instead of relying on untrusted Origin/Referer headers, the system has switched to the "Open CORS, tighten URL validation" model, allowing iframes to be embedded anywhere but only handling requests with valid URLs.

  • FIXED - Unexpected token 'S', ... is not valid JSON: Standardize all API error responses to JSON format. Frontend logic is also enhanced to handle these errors smoothly and display friendly messages to users instead of causing crashes.

  • IMPROVED - Rate Limiting: Implement a rate limiting mechanism for API endpoints, helping to prevent denial of service (DoS) attacks and basic abuse.

  • IMPROVED - Immunity Health Check: Separate the /healthz endpoint from security middleware, ensuring that automated monitoring systems can accurately check application health without being blocked by CORS or other authentication mechanisms.

  • IMPROVED - Content Caching: Add a simple server-side cache mechanism to store crawled post content, which helps reduce the number of requests to the blog page and speed up responses to repeated queries over a period of time.

  • IMPROVED - Project Structure: The source code is organized in a clear structure, separating routes, services, and middleware, making maintenance and future expansion easier.

Deployment and Configuration Guide

  1. Environment Variable Configuration: Create a .env file and provide GEMINI_API_KEY, or you can use 'Environment Varriable' in Render.

  2. Whitelist Configuration: Edit the allowedDomains array in the config.js file to add the blog domains you want to support.

  3. Deployment: The application is ready to deploy on platforms like Render, Heroku via the package.json file.

  4. Integrate into Blogger: Use the provided iframe embed code to add the AI chat button to your blog page.

Version 1.0.0 "Phoenix" lays a solid foundation for the development of future advanced features, such as full website crawling, support for more AI models, and deeper analysis of user interactions.

Assets 2
Loading