OpenRouter's September 2024 App & Agent Rankings reveal Hermes Agent as the top token consumer, followed by Claude Code, Kilo Code, Cline, and DeepSeek Harness, showcasing real-world usage of coding-focused AI agents over model benchmarks.
OpenRouter's September token-usage rankings reveal Hermes Agent leading a top-10 dominated by coding agents like Claude Code, Codex, and DeepSeek Harness, signaling a shift from model-centric benchmarks to real-world agent adoption across 60 tools spanning development, creativity, and automation.
The author details six iterations in building DxC, an agent-native WeChat writing system, revealing that 'AI taste' originates not from surface language patterns but from missing real materials, authorial choices, and accountability; the fix requires material-first drafting, structural editing, independent proofreading, and author confirmation.
CrewAI is a 57k-star open-source multi-agent framework that orchestrates role-based AI agents (researcher, writer, reviewer) via sequential or hierarchical processes, adds tooling, memory, human-in-the-loop, observability, and introduces Flows for deterministic control alongside autonomous Crews.
This article analyzes why default multi-agent architectures leak state in long conversations, advocating for single-loop agents with dynamically loaded skills based on usage frequency, using Anthropic's commerce-agents reference implementation to illustrate caching-aware design, handoff vs. delegation distinctions, and evaluation strategies.
This article outlines five common pitfalls in AI skill development — missing acceptance criteria, skipping domain knowledge, misusing LLMs for deterministic tasks, over-automating workflows, and neglecting output examples — with practical fixes like validation checklists, reference materials, script delegation, human checkpoints, and sample-driven formatting.
The awesome-llm-apps GitHub repository offers 100+ end-to-end tested, CI-gated AI applications across 12 categories—from starter agents to multi-agent systems—compatible with major LLMs and licensed Apache-2.0, providing a graded learning path and a testbed for AI agent security research.
This article explores AI agent memory mechanisms, detailing four memory types—in-context, external, episodic, and parametric—with Python implementation examples using ChromaDB and OpenAI embeddings, plus memory management strategies like time-based decay, importance scoring, and consolidation.
This article explains the architecture of large language model applications, distinguishing between simple chat, fixed workflows, autonomous agents, the Model Context Protocol (MCP) for tool integration, and reusable skills, providing a decision framework for choosing the right approach based on task complexity and stability requirements.
This article explains why embedding model selection must precede reranker evaluation in RAG systems, detailing a three-stage evaluation methodology using business-specific data to measure recall, ranking quality, and end-to-end answer validity, while accounting for engineering constraints like latency, resource usage, and failure modes.
This guide presents a comprehensive prompt engineering framework—including the CRIS four‑step template, RAG‑based prompt construction, and Agent‑oriented architectures—illustrated with practical examples and optimization tips for tasks such as code generation, data extraction, and customer support, helping developers achieve stable, accurate LLM results.
The article argues AI expands software capabilities for previously unsolvable problems but won't replace deterministic systems because accumulated business knowledge, reliability requirements, and ROI make rewriting unjustified; code agents generate deterministic code as a pragmatic compromise.
The author reflects on Manus's acquisition by Meta, arguing that the real value in AI lies not in training foundation models but in engineering the last mile—memory systems, tool sandboxes, and agent frameworks—that turn non-deterministic LLMs into reliable products, while analyzing five business models for AI companies.
This article introduces the fundamental paradigms of large language models, explaining how simple prompts, messages, and tools like RAG and ReAct enable powerful applications, while providing practical code examples, translation strategies, and insights on prompt engineering, tool usage, and model fine‑tuning.
This article explains what Prompt Engineering is, traces its development from early command‑based interactions to modern adaptive and multimodal prompting, details various prompting techniques such as zero‑shot, few‑shot, Chain‑of‑Thought, hallucination‑reduction methods, and demonstrates their practical use in a JD Logistics SKU piece‑type classification case with code examples.
This article explores fifteen detailed examples of how major companies across sectors—from streaming and e‑commerce to transportation and social platforms—are harnessing large language models to improve search, personalize communications, detect fraud, and enhance operational efficiency.
The article examines two emerging AI programming paradigms—visual, node‑based development with ComfyUI for image generation and modular LLM application construction with LangChain—detailing their architectures, key components, workflow examples, advantages, limitations, and practical guidance for leveraging these tools to boost development efficiency in the rapidly evolving AI landscape.
The paper describes Baidu’s AI‑native advertising platform that employs a multi‑agent architecture built on large‑language models—combining large‑small model collaboration, domain SOP‑driven coordination, and long‑term memory—to enable natural‑language understanding, proactive planning, execution and human‑like responses, illustrated by GBI analytics and JarvisBot operations, delivering higher consumption, accuracy, speed and efficiency.
This article examines how large language models such as GPT and Transformer‑based architectures are reshaping the automotive sector by enhancing in‑vehicle intelligence, streamlining product development, improving customer service, and redefining data analyst roles, while also presenting practical experiments, deployment challenges, and future directions.
DingoDB is a multi‑modal vector database that unifies storage and analysis of structured, semi‑structured and unstructured data through a Raft‑based distributed architecture, offering MySQL‑compatible SQL, high‑performance APIs, automatic sharding, real‑time index optimization, and hybrid scalar‑vector queries for enterprise knowledge bases, LLM memory, and real‑time decision‑making.
At CNCC 2023, leading researchers and industry experts convened to examine how large language models can transform recommendation systems, outlining four core challenges—model integration, fluency versus intelligence, hallucination versus deception, and user understanding—while highlighting opportunities such as multimodal content, cold‑start solutions, zero‑shot ranking, instruction‑driven algorithms, and responsible, interactive recommendation pipelines.
This article provides programmers with a comprehensive overview of large language models—including their evolution, core concepts, data pipelines, model architectures, training techniques such as 3D parallelism, supervised fine‑tuning, RLHF, open‑source recipes, and emerging application ecosystems—while also highlighting current challenges and future directions.
