github
2026-03-05
8.5/10
Show summary
Anthropic's official public repository for Claude Agent Skills gained nearly 8,000 stars this week, reaching 84,579 total stars, representing Anthropic's framework for modular, reusable agent capabilities.
github
2026-03-05
8.5/10
Show summary
ByteDance's open-source SuperAgent framework using sandboxes, memory, tools, skills, and subagents to handle long-horizon tasks lasting minutes to hours. Gained 4,136 stars in one week.
github
2026-03-05
8.0/10
Show summary
Open-source extensible AI agent built in Rust that goes beyond code suggestions to install, execute, edit, and test with any LLM backend. Has 32K+ stars with strong weekly growth.
producthunt
2026-03-05
8/10
Show summary
Anthropic launches Claude Opus 4.6, its most advanced model optimized for agentic tasks, with high community traction on ProductHunt. Marks a significant capability step for autonomous agent workflows.
github
2026-03-05
7.5/10
Show summary
Mobile-Agent is a GUI agent family for mobile device control, gaining 649 stars this week with nearly 8,000 total stars.
github
2026-03-05
7.5/10
Show summary
Alibaba's OpenSandbox is a general-purpose sandbox platform for AI applications supporting coding agents, GUI agents, RL training, and code execution with Docker/Kubernetes runtimes, gaining 4,592 stars this week.
github
2026-03-05
7.5/10
Show summary
Self-hostable AI second brain that supports web/doc search, custom agents, automations, and deep research across GPT, Claude, Gemini, Llama, and more. 33K+ stars.
trendshift
2026-03-05
7.0/10
Show summary
Flowise is a highly popular visual AI agent builder with 50,000 stars, offering a no-code interface for constructing LLM-powered workflows and agents.
github
2026-03-05
7.0/10
Show summary
AI-native proxy and data plane for agentic apps built in Rust, providing built-in orchestration, safety, observability, and smart LLM routing. Targets production agentic infrastructure.
trendshift
2026-03-05
7.0/10
Show summary
OpenAI's Symphony platform isolates project work into autonomous implementation runs, enabling teams to manage rather than supervise coding agents — a notable shift toward async agentic development.
hackernews
2026-03-05
7/10
Show summary
An MCP server implementation that claims to reduce Claude Code's context consumption by 98% through a context management mode, addressing a major pain point in long coding sessions.
producthunt
2026-03-05
7/10
Show summary
Anthropic's Sonnet 4.6 launches as the most capable model in the Sonnet series to date, complementing the Opus 4.6 release with strong community interest.
producthunt
2026-03-05
7/10
Show summary
Google launches Gemini 3.1 Pro, a new model iteration positioned for complex reasoning and software engineering tasks, competing in the frontier model space.
producthunt
2026-03-05
7/10
Show summary
Happycapy is an 'agent-native computer' platform designed from the ground up for AI agent workflows, topping ProductHunt charts with 1,334 votes and 588 comments. Represents a new category of AI-first computing environments.
github
2026-03-05
6.5/10
Show summary
A curated collection of LLM app examples using AI agents and RAG across OpenAI, Anthropic, Gemini, and open-source models, approaching 100K stars with 2,647 new stars this week.
trendshift
2026-03-05
6.5/10
Show summary
AgentScope is a framework for building transparent and auditable AI agents with 17,000 stars, emphasizing interpretability and trustworthiness.
github
2026-03-05
6.5/10
Show summary
Chinese-language educational tutorial series on building AI agents from scratch, covering principles and practical implementation. 25K+ stars with 3,290 gained this week.
github
2026-03-05
6.5/10
Show summary
Comprehensive library of agent skills focused on context engineering, multi-agent architectures, and production agent systems with strong traction (13K+ stars, 2.8K this week).
github
2026-03-05
6.5/10
Show summary
Agent orchestration platform for Claude featuring multi-agent swarms, distributed coordination, RAG integration, and native Claude Code/Codex support; gaining significant traction (4.2K stars this week).
trendshift
2026-03-05
6.0/10
Show summary
A curated set of Claude agent skills for research, science, engineering, analysis, finance, and writing tasks, with 11,900 stars.
trendshift
2026-03-05
6.0/10
Show summary
ReMe is a memory management toolkit for AI agents, providing structured tools to store, recall, and refine agent memory across sessions.
trendshift
2026-03-05
6.0/10
Show summary
Agentic skills framework and software development methodology with very high star count (69K), providing reusable agent capabilities for coding workflows.
trendshift
2026-03-05
6.0/10
Show summary
Curated collection of 900+ agentic skills for Claude Code, Cursor, and other coding agents including official Anthropic and Vercel skills; practical resource for agent builders.
hackernews
2026-03-05
6/10
Show summary
Donald Knuth's paper analyzing Claude's behavior on algorithmic/mathematical problems, examining patterns in how the model approaches cycles and iterative computation. Notable for its author and the rigorous lens applied to LLM reasoning.
producthunt
2026-03-05
6/10
Show summary
Claude adds an 'Import Memory' feature allowing users to migrate conversation history and preferences from ChatGPT, lowering switching friction between frontier AI assistants.
producthunt
2026-03-05
6/10
Show summary
Anything API automatically wraps any website with a structured API interface, likely using AI-powered web scraping/parsing to enable programmatic access to arbitrary web data for agent workflows.
hackernews
2026-03-05
5.5/10
Show summary
Anthropic launches a memory import feature for Claude, allowing users to migrate their conversation history and memory from other AI assistants without starting fresh.
trendshift
2026-03-05
5.5/10
Show summary
A curated registry of 5,400+ OpenClaw agent skills, categorized and filterable, accumulating 26,400 stars.
producthunt
2026-03-05
5.5/10
Show summary
Superset allows running multiple AI coding agents (Claude Code, Codex, etc.) in parallel on a local machine, enabling orchestrated multi-agent development workflows.
arxiv
2026-03-05
5/10
Show summary
τ-Knowledge introduces a realistic agentic benchmark for fintech customer support requiring navigation of ~700 interconnected documents; even frontier models achieve only ~25.5% pass rate.
trendshift
2026-03-05
5.0/10
Show summary
ClawRouter is an agent-native LLM router supporting 41+ models with sub-millisecond routing latency and crypto payment integration via USDC on Base and Solana.
trendshift
2026-03-05
5.0/10
Show summary
Self-hosted AI companion system with realtime voice chat and game-playing capabilities (Minecraft, Factorio), targeting Neuro-sama-like functionality across web/macOS/Windows.
github
2026-03-05
5/10
Show summary
Vercel's unified TypeScript SDK for building AI chat bots across Slack, Teams, Google Chat, Discord, and other platforms. Provides cross-platform abstraction for deploying LLM-powered bots.
hackernews
2026-03-05
5/10
Show summary
A practical case study of using LLMs to analyze terabytes of CI logs via SQL generation, finding LLMs surprisingly effective at translating natural language questions to structured queries at scale.
arxiv
2026-03-05
4.5/10
Show summary
Winning SemEval 2026 system combining graph-based retrieval, reflective prompt evolution, and consistency enforcement for abductive event reasoning, achieving 0.95 accuracy while identifying systematic causal reasoning biases across 14 LLMs.
trendshift
2026-03-05
4.5/10
Show summary
Collection of specialized AI agents with defined personalities and workflows for agency-style work including frontend dev, content, and QA roles.
trendshift
2026-03-05
4.5/10
Show summary
High-performance browser automation bridge with multi-instance orchestration and stealth injection, useful as an agent tool for web interaction.
github_developers
2026-03-05
4/10
Show summary
CowAgent is a multi-platform AI assistant supporting proactive task planning, system/resource access, skill creation, long-term memory, and integrations with Feishu, DingTalk, WeChat, etc., supporting OpenAI/Claude/Gemini/DeepSeek/Qwen backends.
arxiv
2026-03-05
3/10
Show summary
Proposes a dual-helix governance framework using a knowledge graph substrate for agentic AI in WebGIS development, reducing cyclomatic complexity by 51% vs. zero-shot LLM approach.
github_developers
2026-03-05
3.0/10
Show summary
Developer profile trending for Squad, a tool for creating AI agent teams for projects.
hackernews
2026-03-05
3/10
Show summary
OpenAI teases an upcoming GPT-5.4 release. Minimal technical detail available.
producthunt
2026-03-05
3/10
Show summary
Lovon is an AI therapy chatbot for emotional support and mental health conversations. Consumer AI application with high ProductHunt engagement but limited technical novelty.
arxiv
2026-03-05
2.5/10
Show summary
Formulates delay-constrained network routing as a constrained MDP and applies constrained deep RL to minimize resource cost while maintaining packet delivery reliability above a target threshold.
github_developers
2026-03-05
2.0/10
Show summary
Developer profile focused on coding agent terminal session management via tmux and git worktrees for tools like Claude Code and Gemini CLI.
hackernews
2026-03-05
2/10
Show summary
User observation that GPT-5.3 makes fear-driven/warning-heavy follow-up prompt suggestions compared to earlier models. Subjective behavioral observation with minimal technical depth.
github_developers
2026-03-05
1.5/10
Show summary
Trending developer sharing personal coding agent configuration.