David-patrick-chuks/Riona-AI-Agent
Riona AI Agent
Riona-AI-Agent 是一个 LLM 自动化代理和浏览器自动化框架，旨在执行复杂工作流、基于自定义数据进行训练并生成 AI 内容。它作为一个社交媒体自动化工具，用于安排内容、管理多个配置文件并自动化跨社交平台的互动操作。
该系统包括一个用于跟踪 AI 代理实时状态、运行时摘要和活动日志的集中式监控仪表板。它利用机器学习模型在自动化浏览会话期间识别并绕过视觉安全挑战。
该代理的能力涵盖社交媒体自动化，包括重复互动的自动化和帖子排期。它支持通过摄取网页链接、音频和文档来训练 AI 代理，以自定义代理的知识库和行为。
该框架使用 Node.js 和 TypeScript 构建。
Features
- Agentic Workflow Automation - Provides an intelligent agent that coordinates complex workflows, trains on custom data, and generates AI content.
- Agent Knowledge Customization - Allows training agents by ingesting information from web links, audio files, and documents.
- Agent Training - Customizes AI agent knowledge and behavior by ingesting external web links and documents.
- Social Media Automation - Automates repetitive interactions and schedules content across multiple social platforms to increase engagement.
- Social Posting Schedulers - Queues images and text for automated publication at specific future times across social platforms.
- Style-Guided Generation - Generates AI captions and comments that adhere to pre-defined styles and brand voices.
- AI Content Generators - Creates AI-generated captions and comments based on specific formats and guided responses.
- Social Engagement Automations - Automates repetitive social interactions such as liking, commenting, and messaging to increase account engagement.
- Knowledge Base Ingestion - Ingests web links, audio, and documents to build a context-aware knowledge base for agent behavior.
- Browser Automation Frameworks - Implements a Node.js and TypeScript framework for programmatically automating web browser interactions and solving captchas.
- Account Profile Switching - Ships a mechanism to switch between multiple authenticated social media profiles using a central configuration file.
- Social Media Automation Clients - Offers a unified interface to automate social media actions including posting, session management, and engagement.
- Asynchronous Agent Job Execution - Manages and triggers long-running AI agent tasks asynchronously to handle complex operational logic.
- CAPTCHA Solving - Uses machine learning to solve CAPTCHA challenges during automated browsing sessions.
- Automated Captcha Solvers - Integrates machine learning models to identify and bypass visual security challenges during automated browsing.
- Multi-Account Session Management - Enables switching and persisting multiple authenticated sessions for different social media profiles.
- Agent Health Monitoring - Tracks the live status, connectivity, and runtime health of autonomous AI agents.
- AI Agent Execution Monitors - Ships a centralized dashboard for capturing execution events and monitoring real-time AI agent performance.
- Automation Activity Logging - Records detailed execution logs and operation summaries of automated agent actions for auditing.
- Centralized Logging Systems - Implements a centralized system for aggregating and storing runtime events and operation summaries for auditing.
- Real-Time Monitoring Dashboards - Provides a dedicated dashboard for tracking the live status and recent action histories of AI agents.
