Source: Post by Rakesh Gohel on Linkedin AI Agent trends have drastically changed from 2024 to 2025 Here are the new trends driving innovation this year.... The Agentic AI field is constantly moving forward with new innovations and products. I'll share with you a few notable trends that are driving the market in 2025. Note: I've also included a few Langgraph codebooks if you want to build some of these solutions yourself. 📌 Agentic RAG - AI Agent workflow for reasoning-based real-time data retrieval and generation. - Agentic RAG is not limited to one use case but rather used in health care as well. - Used by companies like Perplexity, Harvey AI and Glean AI. 📌 Voice Agents - Intelligent agents that interact with users through natural spoken language utlizing a wide range of TTS and STTS embedding and Retrieval. - Used by companies like ElevenLabs, Cognigy, Vapi and Deepgram. 📌 AI Agent Protocols - Streamlining Multi-Agent Communication that supports communication between Agents built in different frameworks. - Used by Companies like Accenture, the top players include A2A, ACP, SLIM and others. 📌 CUA(Computer Using Agents) - AI agents that can interact with a computer the way a human does utilizing tools like a browser, CLI and even a mouse's cursor. - Used by Companies like OpenAI for their Operator, Claude's Computer Use, Runner H by H-Company and Manus AI. 📌 Coding Agents - Multi-Agents for building and debugging applications 10x faster with the help of clever tool use and LLM-based code gen. - Used by companies like Windsurf, Cursor and GitHub Copilot. 📌 Deepresearch Agents - Collaborative Multi-Agent system to build extensively researched reports from a large number of sources. - Popular products include Gemini DR, OpenAI DR and You(.)com DR. 🔗 Few Langgraph Codebooks to build some of the trending solutions 1. Agentic RAG: https://lnkd.in/gdqWyP3r 2. CUA: https://lnkd.in/gCdxpUBi 3. Deepresearch: https://lnkd.in/gexsW5qh 4. Code Gen Agents using RAG: https://lnkd.in/gDnsBFuM 📌 After months of feedback and iteration, we are finally releasing our first technical cohort, "AI Agent Engineering" 🔗 Enrol here: https://lnkd.in/gDEPcXBB If you are a business leader, we've developed frameworks that cut through the hype, including our five-level Agentic AI Progression Framework to evaluate any agent's capabilities in my latest book. 🔗 Book info: https://amzn.to/4irx6nI Save 💾 ➞ React 👍 ➞ Share ♻️ & follow for everything related to AI Agents https://www.linkedin.com/posts/rakeshgohel01_ai-agent-trends-have-drastically-changed-activity-7343981787842297858-PJkS #agenticrag #voiceagents #aiagentprotocols #cua #codingagents #deepresearchagents #langgraph #elevenlabs #cognigy #perplexity #harveyai #gleanai #openai #cursor #githubcopilot #geminidr #youcom #windsurf #a2a #acp #slim #tts #sts #rag #multiagent #aiagents #llm #browserautomation #aicoding

PLINY THE PROMPTER Discusses various advancements in the field of autonomous red teaming, specifically focusing on jailbreak techniques for language models. It highlights the contributions of a prominent figure, Pliny the Prompter, in developing effective jailbreak prompts and attack strategies. Additionally, it addresses ongoing research aimed at enhancing defenses against these vulnerabilities, emphasizing the importance of understanding and mitigating jailbreak risks through comprehensive studies and innovative methodologies. Key Points The document introduces "AutoRedTeamer," emphasizing its capacity for lifelong attack integration in red teaming. "Pliny the Prompter" is credited with devising a highly effective jailbreak prompt that deepens the understanding of language model vulnerabilities. The L1B3RT4S project demonstrates manual attack methods using leetspeak encoding, contributing to broader jailbreak techniques. Current research on bijection learning attacks presents competitive alternatives to established jailbreak methods pioneered by Pliny. The "DeepSeek-R1" project illustrates how behavior modification can be tailored through mixtures of tunable experts, drawing on existing jailbreak strategies. Research on constitutional classifiers is focused on defending against universal jailbreaks by leveraging insights from extensive red teaming exercises. The RoboPAIR platform investigates jailbreaking within LLM-controlled robotic systems, expanding the application of prompt-based attacks beyond traditional language models. https://pliny.gg/ #PlinyThePrompter #AutoRedTeamer #L1B3RT4S #DeepSeekR1 #RoboPAIR #JailbreakLLM #AISecurity #RedTeaming #PromptEngineering #LanguageModels #LLMVulnerability #AIJailbreak #ConstitutionalAI #AdversarialAI #BijectionLearning #AISafety #LLMSecurity #AIResearch

Anthropic Academy - Build with Claude Start developing Claude-powered applications with our comprehensive API guides and best practices https://www.anthropic.com/learn/build-with-claude What you learn (Explore by topic) - Claude 4 Learn how to harness the intelligence and capability of our latest models, Claude Sonnet 4 and Claude Opus 4 - APIs & SDKs Learn about Anthropic's APIs, SDKs, and other development tools - Agents Build autonomous agents and agentic systems that understand, plan, and execute complex tasks - Model Context Protocol (MCP) Build advanced applications with the Model Context Protocol - Claude Code Speed up your development with Claude Code - Tool use Extend Claude's capabilities by connecting to external tools and APIs - Extended thinking Improve Claude's ability to solve complex tasks by allowing it to reason - Retrieval augmented generation (RAG) Build effective RAG systems to enhance Claude's responses with external data - Prompt engineering Create effective prompts that maximize Claude's performance - Evaluations Test and improve Claude's performance with structured evaluations - Prompt caching Optimize performance and reduce costs by reusing Claude's responses - Vision Harness Claude's ability to understand and analyze visual information - Computer use Learn how to use Claude models to interact with a computer desktop environment - Hackathon hacker guide Resources to help you get started with Claude, whether you're hacking with us at an event or on your own

How to use Gemini Code CLI The Gemini CLI is an open-source AI agent that brings the power of Google's Gemini models directly into your terminal. It allows you to interact with Gemini for various tasks, including code understanding, editing, and even automating operational tasks. The CLI uses a "reason and act" (ReAct) loop, leveraging built-in tools and external servers to complete complex requests. To use Gemini Code CLI in VS Code, first install the Gemini Code Assist extension from the VS Code Marketplace. Then, you can interact with it by opening the command palette (Ctrl/Cmd + Shift + P) and typing Gemini: Start Chat to begin a chat session, or using the /generate command in a code file to generate code based on your prompts. https://blog.google/technology/developers/introducing-gemini-cli-open-source-ai-agent/ #GeminiCLI #GeminiCodeAssist #AIagent #opensource #googleai #vscode #codegeneration #ReAct #terminal #commandline #codeunderstanding #codeediting #automation #codepilot #githubcopilot #amazoncodewhisperer #aicoding #googlecloud #developers #codingtools

Abacus.AI has ChatLLM and AppLLM. ChatLLM is an AI assistant for various tasks like chatting, coding, and image generation, while AppLLM is a no-code platform that builds websites and apps from text prompts using a concept called "vibe coding". Essentially, ChatLLM is for conversational AI and general tasks, and AppLLM is for building applications with AI. Here's a breakdown: ChatLLM: Aims to be a comprehensive AI assistant, integrating various AI models for tasks like chatting, coding, and image generation. Features include voice mode, text-to-image generation, document summarization, and code debugging. Can be used for tasks like brainstorming, writing, research, and more. AppLLM: Focuses on building websites and applications from simple text prompts using "vibe coding". Allows users to create functional apps without coding or server setup. Integrates with tools like ChatLLM and CodeLLM for a streamlined AI development workflow. Can build various types of applications, including websites, chatbots, and more complex AI-powered applications. In short, ChatLLM is about interacting with and leveraging AI for diverse tasks, while AppLLM is about using AI to build applications, particularly websites, without traditional coding. https://appllm.abacus.ai/login #AbacusAI #ChatLLM #AppLLM #NoCode #AICoding #VibeCoding #WebsiteBuilder #AppBuilder #AIAssistant #Chatbot #TextToImage #OpenAI #Bubbleio #Adalo

How l switched back to Augment Code from Claude Code ! Tried new Claude Code ... verdict: Worth every penny love the little to-do lists but hell lot expensive ! So what do l do l head back home to the loving arms of Augment Code ... still the best and always will be ! (For now 🤫 ... you know relationships nowadays) So this article exaplins why they are nota big fan of Model Pickers ... makes sense though would be great to have the ooption still !). Here is a low done on Augment code and its recent updates: Augment Code is considered a top AI coding agent due to its advanced context engine, which enables personalized and efficient code generation, and its seamless integration with popular IDEs. It also boasts features like memory persistence, which allows it to adapt to individual coding styles, and Multi-Context Programming (MCP) for connecting to various tools and systems, enhancing its utility in the development workflow. Here's a more detailed breakdown of why Augment Code stands out: 1. Contextual Awareness and Memory: Context Engine: Augment's Context Engine is a key differentiator. It analyzes the entire codebase in real-time, providing context-aware suggestions and code completions, leading to more accurate and relevant AI-driven code generation. Memory Persistence: The platform remembers your coding style and project patterns, ensuring that its suggestions are tailored to your preferences and codebase over time, improving efficiency and reducing errors. MCP (Multi-Context Programming): Augment Code goes beyond just the code by integrating with various tools and systems, such as Vercel, Cloudflare, and more, allowing it to gather more information, automate tasks, and even fix issues in live systems. 2. IDE Integration and Workflow: Seamless Integration: Unlike some AI coding tools that require a separate editor, Augment Code integrates directly with popular IDEs like VS Code, JetBrains, Vim, and GitHub, allowing developers to leverage its AI capabilities within their familiar environment. Code Checkpoints: Augment Code automatically tracks changes and creates checkpoints, enabling easy rollback and providing peace of mind when the agent tackles complex tasks. Remote Agent Functionality: Augment Code can operate in a separate container, allowing developers to work on code even when their main machine is off or unavailable, and even run multiple agents in parallel. 3. Advanced Features and Capabilities: Multi-Modal Support: Augment Code can handle various inputs, including screenshots and Figma files, making it helpful for implementing UI elements and debugging visual issues. Terminal Interaction: Beyond code editing, Augment Code can run terminal commands, streamlining tasks like installing dependencies or running dev servers. Auto Mode: For a more streamlined experience, Augment Code offers an Auto Mode, where it automatically applies suggested changes without requiring explicit confirmation for each action. 4. Focus on Collaboration and Productivity: Developer-Centric: Augment Code is designed to work alongside developers, enhancing their existing workflow rather than replacing it, making it a collaborative rather than a replacement tool. Time Savings: By automating tasks, providing accurate suggestions, and handling repetitive coding tasks, Augment Code frees up developers' time to focus on more complex and creative aspects of their work. Continuous Improvement: Through feedback loops and learning from user interactions, Augment Code continuously improves its performance and adapts to the evolving needs of developers. https://www.augmentcode.com/blog/ai-model-pickers-are-a-design-failure-not-a-feature #augmentcode #claude #aicodingagent #aicode #vscode #jetbrains #vim #github #ideintegration #codingtools #productivitytools #memorypersistence #contextengine #multicontextprogramming #vercel #cloudflare #codecheckpoints #remotework #multimodal #figma #terminalinteraction #automode #developercentric #aitools #aidesign #coding #softwaredevelopment #aitool

RA.Aid (pronounced "raid") helps you develop software autonomously. It is a standalone coding agent built on LangGraph's agent-based task execution framework. The tool provides an intelligent assistant that can help with research, planning, and implementation of multi-step development tasks. RA.Aid can optionally integrate with aider (https://aider.chat/) via the --use-aider flag to leverage its specialized code editing capabilities. The result is near-fully-autonomous software development. https://www.ra-aid.ai/ #RAAid #LangGraph #Aider #autonomouscoding #codingagent #aidev #airesearch #aiplanning #aiimplementation #softwaredevelopment #aidevelopment #autonomoussoftware #aiderchat

Autonomous Design Agent: Lovart Loveart is a revolutionary AI design agent that acts as a fully autonomous design team, capable of handling everything from brand strategy to 3D models and animations, all from a single text prompt. It essentially transforms the creative process, offering agencies and individuals a powerful tool that can analyze briefs, research references, make design decisions, and produce production-ready assets in a fraction of the time and cost of traditional methods. Lovart is described as the world's first fully autonomous design agent, an AI that functions as a 24/7 design team. It transforms prompts into visual content, from storyboards to brand visuals, using auto-design technology. An AI design agent generally assists in creating, editing, and managing visual projects leveraging artificial intelligence. LoveArt appears to offer a comprehensive design solution https://www.lovart.ai/

No Code App Builder: Databutton Databutton is an AI app builder that simplifies the deployment process, enabling users to deploy applications to their own domains quickly. It supports deployment to AWS and Google Cloud. The platform caters to both technical and non-technical users, including founders building SaaS products. Databutton is designed for creating prototypes and well-functioning applications. #databutton #nocode #ai #appbuilder #aws #googlecloud #deployment #saas #prototyping #lowcode #streamlit #bubbleio #appsheet #draganddrop #nocodeappdevelopment #serverless https://databutton.com/ https://youtu.be/T1W-GU6btDY

Check this Repo out ! Shubham Saboo's GitHub repository hosts a collection of LLM Apps. These applications leverage large language models (LLMs) from various providers, including OpenAI, Anthropic, and Google, and open-source models. The apps incorporate Retrieval-Augmented Generation (RAG), AI agents, multi-agent teams, and voice agent technologies. This curated list is a valuable resource for exploring the capabilities and applications of LLMs in different contexts. #AwesomeLLMApps #LLM #LargeLanguageModels #OpenAI #Anthropic #GoogleAI #RAG #RetrievalAugmentedGeneration #AIAgents #MultiAgentTeams #VoiceAgents #GitHub #MachineLearning #NLP #AI #Langchain #LlamaIndex #DeepLearning #AISolutions https://github.com/Shubhamsaboo/awesome-llm-apps

Who has competition now ??? Perplexity ! Tool: Scira AI is a minimalistic, AI-powered search engine designed to help users find information online. It was formerly known as MiniPerplx and provides citations for its search results. Its key function is to offer a streamlined search experience by leveraging AI to deliver relevant information efficiently. Scira's development is powered by Vercel AI. https://scira.ai/ https://youtu.be/z3auANyBk0U?si=awutMT49-mYV97hd https://www.producthunt.com/products/scira https://github.com/zaidmukaddam/scira #SciraAI #MiniPerplx #PerplexityAI #VercelAI #AISearchEngine #SearchEngine #AI #MinimalistSearch #CitationBasedSearch #InformationRetrieval #WebSearch #AlternativeSearch #AISolutions #SemanticSearch

Tool: Qodo AI is a platform focused on improving code quality and developer workflows through AI-powered agents. It offers tools for code generation, testing, and review within integrated development environments (IDEs) and Git workflows. Qodo aims to help developers write better code, identify and fix bugs, and streamline the code review process. So this was Codium before in case wandering so more like a rebrand. Qodo is a quality-first generative AI coding platform, offering developers tools for writing, testing, and reviewing code. Using Qodo, developers can leverage the power of AI directly within their IDE and Git, ensuring that generated code is accurate and high quality. https://www.youtube.com/shorts/IYORBrFyjHM https://www.qodo.ai/ Other Links: https://www.youtube.com/@QodoAI https://youtu.be/-aSX19T6TaY?si=Vw5BMVa95HNry2NU https://www.linkedin.com/posts/qodoai_introducing-qodo-merge-10-to-tackle-ai-assisted-activity-7290368173818761216-2piy/ #QodoAI #CodiumAI #AIcoding #codegeneration #codetesting #codereview #IDE #Git #developerworkflow #AIagents #codequality #jetbrains #visualstudiocode #githubcopilot #amazoncodewhisperer #AICodeAssistant #generativeAI #softwaredevelopment