Best of ai-agents — July 2025

1
Article
Hacker News·44w
Open-Source Agentic Browser
BrowserOS is an open-source browser built on Chromium that integrates local AI agents for automating web tasks. It emphasizes privacy by running AI models locally using Ollama, avoiding data collection by search and ad companies. The browser includes features like automated workflow execution, semantic search over browsing history, and an LLM-based ad-blocker. Compatible with existing Chrome extensions, it targets users seeking privacy-focused browsing with AI-powered productivity tools.
196
13
2
Article
LangChain·44w
How to Build an Agent
A comprehensive framework for building AI agents from concept to production, covering six key steps: defining realistic tasks with concrete examples, creating standard operating procedures, building an MVP with focused prompts, connecting to real data sources, testing and iteration, and deployment with continuous refinement. The guide emphasizes starting small with well-scoped problems, focusing on core LLM reasoning tasks first, and treating deployment as the beginning of iteration rather than the end of development.
163
3
Article
Docker·45w
Top 5 MCP Server Best Practices
Five essential best practices for building MCP (Model Context Protocol) servers: manage tool budget by avoiding one-tool-per-endpoint patterns, design for AI agents rather than end users with proper error handling, document for both human users and AI agents, test user interactions beyond just functionality using MCP inspector, and package servers as Docker containers for portability. The guide emphasizes that AI agents are the actual consumers of MCP tools, requiring different design considerations than traditional user-facing APIs.
102
4
Article
freeCodeCamp·43w
How AI Agents Remember Things: The Role of Vector Stores in LLM Memory
Large language models don't have inherent memory, but vector stores enable AI agents to simulate memory by converting text into numerical embeddings and storing them in specialized databases. When users interact with AI, the system searches for semantically similar stored vectors to retrieve relevant past information. Popular vector databases include FAISS for local deployments and Pinecone for cloud-based solutions. This approach, called retrieval-augmented generation (RAG), allows AI to appear contextually aware despite technical limitations around similarity-based matching and static embeddings.
98
5
Article
The New Stack·45w
How (Human) Developers Should Upskill in the AI Era
Developers must adapt their skills for an AI-driven future by focusing on agent orchestration, business process understanding, and systems thinking rather than traditional coding. The new full-stack development includes technical depth, business acumen, and data science capabilities. Key areas for upskilling include building AI agent workflows, understanding business domains, and managing non-deterministic AI systems with proper evaluation and governance.
93
9
6
Article
Tinybird·42w
Why LLMs struggle with analytics
LLMs face significant challenges when working with analytical data, struggling with tabular data interpretation, SQL generation accuracy, and complex database schemas. The key to successful agentic analytics lies in providing comprehensive context through detailed documentation, semantic models, and sample data rather than expecting perfect SQL generation. Building query validation loops with error feedback, using LLM-as-a-judge evaluators, and focusing on business understanding over technical perfection enables more reliable analytical insights.
90
5
7
Article
Javarevisited·44w
8 Best Udemy Courses to Learn n8n for AI Automation and AI Agents in 2025
A curated list of 8 Udemy courses for learning n8n, an open-source workflow automation platform for building AI agents and automations. The courses cover topics from beginner-friendly AI automation to advanced RAG-based agents, voice AI, and monetizing automation skills. Each course targets different skill levels and use cases, from non-technical users to developers building complex AI agent architectures.
78
8
Article
Vercel·43w
Grep a million GitHub repositories via MCP
Grep now supports the Model Context Protocol (MCP), allowing AI agents to search over a million public GitHub repositories through a standardized interface. The MCP server can be easily integrated with AI clients like Cursor and Claude, enabling agents to query code patterns and retrieve relevant snippets in real-time. Vercel built the MCP server quickly using their mcp-handler package, which simplifies the process of exposing existing APIs to AI clients.
77
2
9
Article
Javarevisited·41w
Top 5 Udemy Courses to Learn Claude Code and Claude AI in 2025
Claude AI and Claude Code are emerging as powerful tools in the AI development stack, created by Anthropic with a focus on safety and natural language understanding. Claude Code enables developers to write production-ready code through conversational prompts and automate workflows with AI agents. The article curates five Udemy courses covering different aspects: from basic Claude Code usage and full-stack AI development to advanced agent building with frameworks like LangChain, CrewAI, and AutoGen. These courses cater to various skill levels and use cases, from beginners learning AI-assisted coding to experienced developers building complex autonomous agents.
75
1
10
Video
YouTube·44w
How to Build AI Agents with n8n in 2025! (Full Course)
A comprehensive tutorial covering how to build AI agents and automations using n8n, a no-code workflow automation platform. The guide starts with fundamental concepts like the difference between agents and automations, explains n8n's node-based system including triggers, actions, utilities, code nodes, and AI components. It walks through creating a practical lead form automation that processes form submissions and sends labeled email notifications based on project budget. The tutorial emphasizes hands-on learning with step-by-step instructions for setting up workflows, configuring nodes, and understanding data flow between components.
70
11
Article
Daily Dose of Data Science | Avi Chawla | Substack·44w
MCP Integration with 4 Popular Agentic Frameworks
Part 8 of an MCP crash course demonstrates how to integrate Model Context Protocol with four popular agentic frameworks: LangGraph, CrewAI, LlamaIndex, and PydanticAI. The tutorial provides step-by-step practical walkthroughs for connecting MCP to each framework, along with detailed implementations. This builds on previous parts covering MCP fundamentals, custom client development, tools/resources/prompts, sampling integration, and security considerations including testing and sandboxing.
67
12
Article
Daily Dose of Data Science | Avi Chawla | Substack·42w
What is Context Engineering?
Context engineering is emerging as a critical skill for AI engineers, focusing on systematically orchestrating context rather than just clever prompting. Unlike traditional prompt engineering that relies on 'magic words', context engineering creates dynamic systems that provide the right information, tools, and format to LLMs. The approach addresses the real bottleneck in AI applications - not model capability, but setting up proper information architecture. Key components include dynamic information flow, smart tool access, memory management (both short-term and long-term), and format optimization. As AI models improve, context quality becomes the limiting factor for application success.
62
13
Video
YouTube·44w
Complete Guide to Build and Deploy an AI Agent with Docker Containers and Python
A comprehensive guide covering Docker fundamentals and building AI agents with Python. Starts with Docker basics including container creation, image building, and Docker Compose usage. Progresses through setting up FastAPI web applications, integrating databases, and ultimately implementing AI agents using Langchain and Langraph. Covers both local development with Docker containers and deployment strategies using services like Railway and Digital Ocean. Demonstrates how to use both managed LLM services and open-source AI models available through DockerHub.
59
1
14
Article
Daily Dose of Data Science | Avi Chawla | Substack·45w
6 No-code LLM, Agents, and RAG Builder Tools for AI Engineers
Six open-source no-code tools enable AI engineers to build LLM applications, agents, and RAG systems without extensive programming. Featured tools include RAGFlow for document understanding, Langflow for visual agent building, LLaMA-Factory for model fine-tuning, Transformer Lab for local LLM experimentation, xpander for agent backends, and AutoAgent for natural language agent creation. These platforms collectively have over 200k GitHub stars and support various AI development workflows from training to deployment.
40
15
Article
Daily Dose of Data Science | Avi Chawla | Substack·42w
Connect Any LLM to Any MCP server
mcp-use is an open-source library that enables developers to connect any LLM to any MCP (Model Context Protocol) server in just 3 lines of code. Unlike being limited to Claude or Cursor, this tool allows building custom MCP agents with local LLMs like Ollama, supports multiple simultaneous MCP server connections, provides sandboxed execution, and includes debugging capabilities for 100% local MCP client development.
36
16
Article
Daily Dose of Data Science | Avi Chawla | Substack·43w
Build a Multi-agent Content Creation System
A demonstration of building a multi-agent content creation system using Motia, an open-source backend framework that unifies multi-agent orchestration, APIs, and background jobs. The system scrapes web content using Firecrawl, processes it with locally-served Deepseek-R1 LLM through Ollama, and generates social media content for Twitter and LinkedIn in parallel. The workflow includes automatic scheduling via Typefully and exposes functionality through APIs. Motia supports multiple programming languages, one-click deployment, built-in observability, automatic retries, and streaming responses.
36
17
Article
Vercel·41w
AI SDK 5
AI SDK 5 introduces major improvements including type-safe chat integration for React, Vue, Svelte, and Angular with separate UI and model messages for better state management. Key features include agentic loop control with stopWhen and prepareStep functions, experimental speech generation and transcription, enhanced tool capabilities with dynamic tools and provider-executed functions, and full-stack type safety. The release also adds data parts for streaming custom typed data, message metadata support, SSE streaming, and Zod 4 compatibility while maintaining the unified provider API for seamless model switching.
32
1
18
Article
Community Picks·45w
bytedance/trae-agent
Trae Agent is an open-source LLM-powered tool for automating software engineering tasks through natural language commands. It supports multiple LLM providers (OpenAI, Anthropic), offers a CLI interface with interactive mode, and includes built-in tools for file editing, bash execution, and structured problem-solving. The project is in alpha stage and features trajectory recording for debugging, flexible JSON configuration, and plans to migrate to Rust.
23
19
Article
Product Hunt·45w
VoltOps: Trace, debug, and monitor AI agents apps in n8n-style
VoltOps is a developer-first observability platform specifically designed for AI agents and LLM applications. It provides tracing, debugging, and monitoring capabilities for agent workflows with features like structured traces, rich logs, and an n8n-style visual interface. The platform is framework-agnostic and supports multi-step chains, tool calls, and memory operations. It offers JavaScript/TypeScript and Python SDKs, with integrations for VoltAgent and Vercel AI SDK.
24
20
Article
LangChain·43w
Open Deep Research
LangChain introduces an open-source deep research agent built on LangGraph that automates comprehensive research tasks. The system uses a three-phase approach: scoping (clarifying user requirements), research (using supervisor and sub-agents for parallel investigation), and writing (generating final reports). Key insights include using multi-agent architecture only for parallelizable tasks, isolating context across research topics to avoid token bloat, and implementing context engineering to manage computational costs. The agent flexibly adapts research strategies based on request complexity and is available through LangGraph Studio and Open Agent Platform.
22
21
Article
LangChain·41w
Deep Agents
Traditional LLM agents that simply call tools in a loop are limited in handling complex, long-term tasks. Deep agents overcome these limitations through four key components: detailed system prompts with examples, planning tools (like todo lists), sub-agents for task decomposition, and file systems for context management. Applications like Claude Code, Deep Research, and Manus demonstrate this architecture's effectiveness. The author introduces an open-source 'deepagents' package that implements these patterns, making it easier to build specialized deep agents for specific domains.
21
1
22
Article
Hacker News·45w
Stop Building AI Agents
AI agents are often overused and unnecessarily complex for most LLM applications. Instead of jumping straight to agent frameworks, developers should start with simpler workflow patterns like prompt chaining, parallelization, routing, orchestrator-worker, and evaluator-optimizer. These patterns solve most problems more reliably and are easier to debug. Agents work best in human-in-the-loop scenarios where oversight and flexibility are needed, but should be avoided for stable enterprise systems that require deterministic behavior.
19
23
Article
KodeKloud's Squad·43w
MCP (Model Context Protocol) Simplified – But Let's Go Deeper!
Model Context Protocol (MCP) is a shared communication language that enables AI agents to work together in modular, scalable systems. Unlike traditional API gateways, MCP allows context-aware agents to communicate with specialized services and delegate tasks to other agents through protocols like Agent2Agent. This creates composable, decentralized AI systems where multiple specialized agents collaborate rather than relying on a single large model.
17
24
Article
Hacker News·44w
BloopAI/vibe-kanban: Kanban board to manage your AI coding agents
Vibe Kanban is a project management tool designed specifically for orchestrating AI coding agents like Claude Code, Gemini CLI, and Codex. It provides a kanban board interface to manage multiple AI agents working in parallel or sequence, track task status, review code, and centralize configuration. The tool addresses the shift in software development where engineers increasingly focus on planning and orchestrating AI agents rather than writing code directly.
17
25
Article
Fermyon·44w
Serverless A2A with Spin
A2A (Agent-to-Agent) is a new open protocol that enables AI agents to discover and communicate with each other. The protocol uses JSON-based agent cards for discovery and JSON-RPC for interaction, supporting three communication modes: asynchronous sessions, streamed sessions, and push notifications. The tutorial demonstrates building a serverless A2A agent using Spin and WebAssembly that provides ethical reasoning using Google's Gemini model. The agent exposes its capabilities through a standardized agent card and handles requests via JSON-RPC, making it discoverable and interoperable with other A2A-compliant agents.
16

See all ai-agents archives