5/19/2026
Show HN: RTFRA - A Humble Proposal [RFC]
RTFRA proposes AI agents should read documentation. RFC-style proposal about improving documentation standards for AI agent compatibility.
5/19/2026
Show HN: claude-autopilot, autonomous dev pipeline with multi-model review
Open-source autonomous development pipeline for Claude enabling brainstorm-to-merge workflow on local codebases with multi-model review.
5/19/2026
RTK: Rust Token Killer
RTK is an open-source CLI tool that reduces LLM token consumption 60-90% by filtering and compressing command outputs. Single Rust binary supporting 100+ commands.
5/19/2026
Show HN: We built an AI strategy agent that runs consulting-style workflows
AI strategy agent that automates consulting-style workflows for strategic analysis. Addresses gap between expensive consulting and ChatGPT by providing structured analysis.
5/19/2026
Couldn't find privacy filter for Claude, so I built one
Privacy-focused workspace for Claude, ChatGPT, and custom LLMs with policy enforcement, secret protection, sandboxed execution, and multi-provider management via Outgate.
5/19/2026
Show HN: Gutenberg – Any URL to verified CLI and MCP server and agent skills
Tool factory converting any public API into verified CLI, MCP server, and agent skills with caching and proof artifacts.
5/19/2026
Stop babysitting your AI agents
AgentRail connects coding agents to GitHub, Linear, and CircleCI to handle full workflow loops beyond code generation. Manages tickets, CI/CD, reviews, and merging.
5/19/2026
Yet Another AI Teammate
Curated directory of AI teammate products including Tencent's WorkBuddy desktop agent for natural language task automation and multi-step workflow planning.
5/19/2026
Build managed agents with the Gemini API
Gemini API launches Managed Agents supporting custom agents with versioned instructions, skills, data files, tool use, and sandboxed code execution in isolated Linux environments.
5/19/2026
Anthropic Announces Managed Agents – what this means for consumers
Analysis of pricing model shifts: GitHub Copilot moves to usage-based billing, Anthropic caps Claude Code, driving users toward self-hosted open-source alternatives.
5/19/2026
Agentic PCB Design Sucks
Open-source registry of PCB components with structured specs designed for AI agents to resolve and filter parts.
5/19/2026
Show HN: AgentWing – make AI agents complete tasks faster
AgentWing adds director layer for parallel multi-agent task execution, decomposing sequential workflows into independent scoped jobs to improve speed.
5/19/2026
Google Antigravity Python SDK
Python SDK for building AI agents with Google Antigravity and Gemini, providing secure stateful infrastructure that abstracts agentic loop mechanics.
5/19/2026
AgentPTY: Script Claude Code’s interactive CLI through a persistent PTY
AgentPTY wraps Claude Code's interactive CLI in persistent pseudo-terminal, exposing scriptable session via FastAPI for programmatic agent interaction and automation.
5/19/2026
Agent First Language by Vercel
Zero: agent-first programming language experiment by Vercel where agents are primary users, featuring on-the-fly learning, structured debugging, and comprehensive standard library.
5/19/2026
Anthropic is killing stainless, so we built our own SDK/MCP generator
Custom SDK/MCP generator tool (Ironic) built to replace Anthropic's Stainless for API client and Model Context Protocol code generation.
5/19/2026
Show HN: I built a native macOS Markdown viewer 100% with AI coding agents
Native macOS Markdown viewer built entirely with AI coding agents using Tauri 2, supporting GitHub Flavored Markdown and Obsidian extensions.
5/19/2026
I carried my stories From Python to Node and found Apple's hidden city of Metal
Local AI model fine-tuning framework for voice cloning built in Python/Node using retrieval-based keyword approach instead of embeddings, enabling voice-consistent text generation.
5/19/2026
Show HN: PrismoDev – local CLI for finding token waste in Claude Code/Codex
PrismoDev CLI tool analyzes Claude Code sessions to identify token waste from context bloat, helping reduce costs from artifacts and logs.
5/19/2026
VeilGate- Deception Reverse Proxy
Deception reverse proxy for AI pentest agents; discusses LLM agent-driven security testing against production systems using MCP servers.
5/19/2026
Show HN: Enforra – open-source action governance for AI agent tool calls
Open-source SDK for AI agent tool call governance, enabling policy enforcement, CI testing, and audit trails before tool execution.
5/19/2026
How-to-Train-Your-GPT
Interactive textbook (12 chapters, 7500+ lines) teaching how to build and train modern language models from scratch, focusing on attention mechanisms.
5/19/2026
Web Researcher MCP: Give AI assistants web search and research capabilities (Go)
Production MCP server enabling AI assistants to perform web search, content extraction, and multi-source research via Model Context Protocol.
5/19/2026
Android Halo helps you stay in sync with your AI agent
Android Halo UI provides real-time visibility into AI agent status and progress notifications on device screens.
5/19/2026
Ask HN: How are agentic workflows meant to offset AI debt?
Discussion on whether agentic workflows can reduce technical debt and improve project design patterns.
5/19/2026
CloudNSite – AI agents that replace manual business processes for SMBs
CloudNSite deploys pre-built AI agents to automate business workflows for SMBs across healthcare, legal, real estate sectors.
5/19/2026
Introducing OpenAI for Singapore
OpenAI partnership with Singapore's Ministry of Digital Development for national AI strategy support. Government initiative announcement.
5/19/2026
Show HN: A macOS menu bar app for monitoring live Claude Code sessions and usage
Native macOS menu bar app monitoring Claude Code sessions and usage limits in real-time.
5/19/2026
Zero: The Programming Language for Agents
Experimental agent-first programming language designed for AI agents as primary users; includes structured debugging tooling and broad standard library.
5/19/2026
Show HN: AgentKit SEO - a career context file that installs as agent skills
Context file system for AI agents that installs as reusable skills. Solves persistent context management across multiple agent sessions for job hunting applications.
5/19/2026
Show HN: YouTube MCP, give any AI agent access to YouTube
Local MCP server enabling Claude, ChatGPT and other AI agents to access YouTube. Provides 8 tools including transcript fetching, summarization, and video metadata extraction.
5/19/2026
Show HN: DDS Vibe Academy – 31 free AI coding masterclasses, built by AI agents
Free 31-class AI coding curriculum covering Claude, Gemini, Cursor, Ollama and others. Academy hub built entirely by AI agents using Claude Opus and Google Antigravity.
5/19/2026
Most AI agent papers stack one LLM with a vector store, we flipped it
Research paper on alternative AI agent architecture using vector store and LLM in reversed configuration.
5/19/2026
The tool that made our AI agent better at using its tools
Tool improvement for AI agents' capability to use tools effectively.
5/19/2026
Ternative – C++/CUDA inference engine for ternary LLMs with runtime LoRA
C++/CUDA inference engine optimized for ternary LLMs with runtime LoRA adaptation support.
5/19/2026
A brief investigation into the GPT-5.5 regression claims
Empirical analysis of GPT-5.5 performance regressions on 21 code tasks from open-source repo; found directional regression on tests but broad quality improvements on code review and craft metrics.
5/19/2026
OpenAI Adopts Google's SynthID Watermark for AI Images with Verification Tool
OpenAI adopts Google's SynthID watermark for AI-generated images with public verification tool.
5/19/2026
The State of Statefulness in AI Agents
Mistral AI acquires Emmi AI to strengthen its AI product stack.
5/19/2026
Tools to understand how content was created and edited
Research article on statefulness in AI agents and maintaining context across interactions.
5/19/2026
Show HN: Logbox – let Claude monitor your dev logs
Open-source logbox tool for Claude to monitor and debug dev server logs via MCP integration.
5/19/2026
The Gemini app becomes more agentic, delivering proactive, 24/7 help
Google Gemini app gains agentic capabilities for proactive 24/7 assistance.
5/19/2026
Categorizing without an LLM
Article about alternative categorization approaches without LLMs. Link-only stub with no detailed content.
5/19/2026
TinyFish Vault: Your Web Agent Can Now Log in Without Touching Your Passwords
TinyFish Vault enables web agents to authenticate securely without direct password access.
5/19/2026
A new era for AI Search
Google Search integrates AI agents enabling complex queries through natural language, with updated search interface.
5/19/2026
Evaluating job search ranking with LLM judged NDCG
Using LLM-judged NDCG metric to evaluate job search ranking quality.
5/19/2026
Intentional Amnesia: Why I Wipe My AI Agents' Memories Every 15 Minutes
Technical architecture for multi-agent orchestration system that periodically wipes AI agent memory to manage state and prevent accumulation.
5/19/2026
Microsoft's Durabletask Package on PyPI Compromised. Mini Shai Hulud
Three malicious versions of durabletask package on PyPI contain dropper malware that executes infostealer payload, targeting developer credentials.
5/19/2026
Show HN: AgentShield – Stop AI agents from spending money unsupervised
AgentShield provides payment safety for autonomous AI agents by enforcing four-check safeguards before financial transactions, preventing budget overruns and prompt injection exploitation.
5/19/2026
Gemini Omni
Google announces Gemini Omni, a natively multimodal model extending video/image generation capabilities with improved reasoning across input types.
5/19/2026
Managed Agents in the Gemini API
Title only, no content available.