Neurix
Reliability layer for AI agents, ensuring production-grade performance.
Vertical Hub
The framework layer of AI moves in waves. This hub tracks AI frameworks gaining real traction: agent runtimes, eval harnesses, RAG toolkits, prompt management layers, tool-calling abstractions. Updated daily from where developers actually congregate — GitHub Trending, HN, the ML subreddits.
Reliability layer for AI agents, ensuring production-grade performance.
Memory layer library for AI agents with long-term recall
Framework for controlling enterprise AI models
Anthropic's new framework for evaluating jailbreak robustness in AI models.
Protocol for secure agent-to-agent interactions
Self-reflective AI system library for agents that debug themselves.
Curated collection of reusable agent loops for AI workflows.
Stateless parallel implementation of x402 for agent payments without sessions.
Dev tool for coordinating multiple specialized AI models in workflows.
Lightweight framework for multi-agent orchestration.
Lightweight agent composition library for multi-agent flows.
Open-source CMS for Next.js and React applications.
New open-source library for building AI agents.
Open-source framework for AI-powered vulnerability discovery
Lightweight agent orchestration library for multi-agent coordination
Playbook for company-wide AI strategy
Self-reflection prompting framework for improved reasoning in open models.
Adaptive training environment for AGI from PatronusAI.
Lightweight framework for non-wrapper AI agents, launched this week.
Library to add voice/video calling capabilities to AI agents.
Python library for building AI agents with structured outputs.
Viral AI agent framework with revolutionary demos.
A new JavaScript runtime and ecosystem, alternative to Node.js.
Optimized runtime for coding agents reducing cost
Geospatial data skill for Claude/Codex AI coding assistants
Tool for managing long-running AI agent prompts with pause/resume.
Verification system proving AI agents executed tasks correctly.
Framework for secure AI execution in Trusted Execution Environments.
AI agent skill for WordPress Playground enabling automated testing and iteration.
Memory layer library for long-term recall in AI agents
Lightweight local agent library gaining rapid GitHub stars.
Curated set of 40 powerful prompts for Claude AI.
Browser-based network traffic inspection and modification proxy.
Adobe Research's AI proxy agents that adapt to designer preferences.
Pure-Rust cryptography library with benchmarks
Debate on platform vs surface for AI-native dev tools
Methodology for AI-reviewed startup applications
GPU-accelerated UI framework for Zig language
SDK for extending Ableton Live
Self-hosted dev sandboxes with preview URLs
Major release of Angular framework
Open-source suite for building defensible AI products.
Multi-agent collaboration methodology for coordinated task execution.
Open-source repo guardrails for reliable AI coding agents.
Universal modular framework for autonomous AI research agents.
LLM framework for automating multimodal neuroimaging analysis without manual scripting.
Methodology for chaining literature insights with agent reasoning.
AI-powered static analysis for code security.
Quick regression testing method for AI agent system prompts.
Framework for multiple AI agents to debate and produce superior outputs.
Tools to chunk, route, and summarize 1M+ token contexts for agents.
Agents with persistent vector/graph memory for long-term tasks.
Apps where core logic lives in model context rather than traditional codebases.
PyTorch library for profiling attention mechanisms in transformers.
Python library implementing Go-style coroutines for free-threaded Python.
Python 3.14 compiled to native code without interpreter
Python library for manipulating SQLite databases, now with AI-assisted code.
Mesh network for distributed FastAPI servers.
Enhanced version of Vite build tool
MCP server for multi-agent structured debates
Joint training framework for next-state generation and actions in world models.
Framework for transferring robot learning from simulation to hardware.
Agent workflow framework for pre-incident mapping and investigation
Open-source Multi-Agent Reinforcement Learning library.
Real-time context layer library for AI apps.
Technical details on smart contracts and bonding curves for AI agents
Microsoft's graph-based RAG methodology for improved knowledge retrieval.
Methodology for building declarative AI systems
Open-source vector database for AI applications
Open-source environment for agentic RL research.
Patterns and tools for coordinating multiple AI agents
Minimalist agent swarm library, compared favorably to heavier frameworks.
AI-native open-source embedding database
NVIDIA's SDK update with agentic AI tools for edge devices.
Framework for stablecoin payments between AI agents by Circle.
End-to-end speech recognition toolkit from ModelScope
Compiler that turns TypeScript into native executables via SWC and LLVM.
Decentralized protocol for multi-agent communication.
Rust-based frontend for vLLM inference engine, improving performance.
LLM runtime that streams models through VRAM from system RAM efficiently.
Distributed inference framework for LLMs across multiple nodes.
Fork of llama.cpp optimized for running large models on limited VRAM.
Speculative decoding framework for faster LLM inference.
Architecture combining diffusion models with transformers for generation.
AI agent that autonomously performs CEO-level strategic planning and decision-making.
Architecture tightly coupling vision, audio, and text in real-time.
Optimized backend for running transformers models with vLLM.
Heterogeneous AI computing management framework
Manticore Search optimized ONNX-based embeddings for faster vector search.
Rust compiler translated entirely to C for portability.
New benchmark for evaluating AI models
Library for spatial awareness in synthetic data for agents.
Game engine update with AI agent integration for game development
Rigorous framework for biology research beyond biases
AI-native education pipeline methodology for dev training
Pattern for using Redis as agent memory layer in production
Pre-mainnet Move Framework upgraded from v22 to v25.
PyTorch library for fusing MLP layers for efficiency.
Open-source UI kit for building document-centric applications.
Rust framework for building cross-platform user interfaces.
Universal server framework for JavaScript/TypeScript.
Improved self-reflective agent loop technique
Modular vision-language architecture from a fresh research paper.
Implementation of Laguna model in llama.cpp for local inference.
Async architectures for managing long-running, multi-session AI agents.
Real-time human-AI collaboration framework for critical decisions
Simple eval to detect regression in AI agent behavior after prompt changes
Clojure foreign function interface for libgdx game development framework
Integration between SkyPilot and Hugging Face storage for ML
Odin programming language 1.0 release
Open-source robotics framework update from Hugging Face
Library for multi-dimensional image similarity
GCC compiler backend targeting WebAssembly.