#AI Agent
1115 related articles
Jev AI Decision Model: 7 Use Cases to Make Your Coding Agent Faster and Cheaper
Jev AI is a fast, low-cost decision model that pairs with coding agents like Claude Code. Explore 7 real-world use cases: context compression, doc validation, skill selection, file ranking, code review, browser testing, and rules enforcement.
Decoding Anthropic's Talk: Models Are Absorbing Your Scaffolding Code
At Code with Claude, Anthropic's Lucas argued that scaffolding developers hand-rolled last year is now built into the model. Covers tool routing, context management, code execution, Opus 4.7's native 1440p support, OS World 78%, and Claude Code tips.
A Browser Agent With 99% Per-Step Accuracy — Why Does It Only Succeed 36% of the Time After 100 Steps?
BrowserBase engineer Derek Megan explains why a 99% per-step browser agent succeeds only 36% of the time over 100 steps — and how engineering fixes it.
AI Digital Employee Systems Decoded: Can Agent Workflows Really Replace Entry-Level Jobs?
A deep dive into AI digital employee systems and Agent workflows — analyzing real capabilities vs. hype around smart agents, digital humans, video matrix, and automated lead gen.
DingTalk's Official Open-Source CLI: How dws Turns Enterprise Collaboration into an AI-Programmable Workbench
DingTalk's official open-source CLI tool dws unifies enterprise collaboration under one entry point, with deep AI Agent support via structured JSON, field projection, OAuth security, and Dry Run execution guards.
Mac Computer Use: The Skill Pack That Lets AI Agents Actually Act on Mac
Mac Computer Use is an open-source skill pack letting AI agents truly execute and verify actions on Mac. Covers observe-act-verify loop, permissions, security boundaries, and compatible clients.
AWS DevOps Agent Auto-Remediation in Practice: From Diagnosis to One-Click Approval
AWS DevOps Agent only diagnoses by default. Learn how Lambda Durable Functions, EventBridge, and Bedrock turn investigation summaries into pre-validated fixes for one-click approval.
Where Does AI Agent Waste Really Come From: Model Calls or the Orchestration Layer?
Is AI Agent waste coming from model calls or the orchestration layer? KORA Doctor finds production waste often stems from repeated tool data fetching, not redundant inference.
OWASP LLM Top 10 (2026): A Complete Guide to AI Agent Security
Based on the OWASP LLM Top 10 (2026) threat map, this guide covers AI Agent security risks — prompt injection, data leakage, code execution, and multi-agent threats — with practical layered defense strategies.
Hands-On Tutorial: Building an AI Agent on TongAgent Platform with MCP Server
Learn how to build an AI agent on TongAgent using MCP Server — covering LLM setup, AMAP MCP tools, template assembly, testing, and publishing. No heavy coding required.
AI Agent Automatically Reviews Microsoft 365 Sign-in and MFA Logs: A Real-World Test
A security engineer tests an AI Agent that autonomously reviews Microsoft 365 sign-in and MFA logs, cross-references official docs, and outputs actionable recommendations.
Text2Dashboard: A Controlled Agent Architecture for Generating Enterprise Dashboards from Natural Language
Text2Dashboard is a controlled Agent architecture prototype for enterprise DataBrain platforms that converts natural language requests into auditable dashboards, separating model proposals from deterministic execution.
Why Do Unattended AI Agents Redo Completed Work? LangGraph's Context Trap
A developer shares how their LangGraph Agent redid completed nodes during unattended runs. Learn why context accumulation — not model drift — is the real culprit, and how active context collapse fixes it.
AI Chatbot Has a Knowledge Base but Still Gets It Wrong? Here's the Real Problem with RAG Retrieval
AI chatbots with knowledge bases still give wrong answers because RAG retrieves semantically similar but condition-mismatched content. Learn how metadata, condition-first retrieval, and conflict arbitration fix this.
Agent Loop Interview Essential: The Fundamental Difference from a Single LLM API Call
Agent Loop is a top interview topic for AI engineers. This article breaks down the 4 key differences from a single LLM API call: lifecycle, tools, memory, and deliverables.
EvoX Agent Tested: How 100 AI Agents Collaborating in a Swarm Pushed Accuracy from 26% to 71%
EvoX Agent tested: a single agent scores 26%, but 100 agents in a swarm hit 71%. We break down swarm collaboration, self-evolving skills, and auto model switching.
AI Agent Automatically Generates BIM Structural Model from 18 CAD/PDF Drawings in 30 Minutes
YouTube channel Next Construct demonstrates an AI Agent building a multi-story Revit BIM structural model from 18 CAD/PDF drawings in ~30 minutes, covering piles, pile caps, 50 columns, and beam-slab systems via MCP.
Are AI Agents Actually Making You More Productive? A Soul-Searching Question from Reddit
A viral Reddit post compares AI Agent productivity to speeding through traffic — are you actually getting anywhere? A deep dive into whether AI Agents deliver real progress or just the illusion of busyness.
Turning APIs into MCP Servers with Cognipeer Console: A Practical Guide to Tool Exposure for AI Agents
Convert OpenAPI specs into MCP servers with Cognipeer Console, making your APIs discoverable and callable by AI Agents. Full walkthrough: server creation, tool inspection, Hub publishing, and monitoring.
Building a Restaurant Reservation AI Agent with n8n: A Telegram Automation Deep Dive
How an n8n + Telegram AI Agent automates restaurant reservations — from natural language requests and conflict detection to manager approval — in just 15 seconds.