3/1/2026
Show HN: Hive Memory – Cross-project memory for AI coding agents (MCP server)
MCP server providing persistent cross-project memory for AI coding agents, maintaining context across project switches.
3/1/2026
Why XML Tags Are So Fundamental to Claude
Article explaining the importance of XML tags in Claude's prompt handling. Minimal content provided.
3/1/2026
Show HN: A web-based vibe coding platform (works from your phone)
Web-based platform for generating working projects from natural language descriptions using Claude Opus, supporting multi-file iteration.
3/1/2026
Handoff: pick up where you left off when switching between Claude Code and Codex
Handoff enables session continuity between Claude Code and Codex CLI. Allows picking up work across different AI coding tools.
3/1/2026
Show HN: MCP Playground – free MCP test servers, inspector, and 10K+ server list
MCP Playground is a Postman-style tool for testing Model Context Protocol servers. Includes browser inspector, public test servers, and 10K+ server directory.
3/1/2026
How HN: Agent-Vault – A Zero-trust credential manager for AI agents
Agent-Vault: zero-trust credential manager for AI agents using local age encryption and Git-based syncing without servers or SaaS.
3/1/2026
Show HN: Tree, but for Token Usage
Treetok analyzes token usage for Claude and OpenAI across directory structures. Useful for understanding context window consumption.
3/1/2026
Claude Dungeon – Visualize Claude Code sessions as pixel-art dungeon heroes
Claude Dungeon visualizes Claude Code agent sessions as pixel-art characters navigating a dungeon. Real-time visualization with React and WebSocket.
3/1/2026
Show HN: Agentic Airport
Browser-based air traffic control simulation using agentic AI for coordinating multiple aircraft landings and collision avoidance.
3/1/2026
Show HN: TheAgentMail – email for AI agents with karma-based spam prevention
TheAgentMail: email API for AI agents with karma-based spam prevention, enabling agents to create temporary email addresses.
3/1/2026
Local LLM compresses long prompts before they reach Claude – MCP server
MCP server implementing semantic compression for LLM prompts, reducing token usage 40-60% while preserving conditional constraints.
3/1/2026
Show HN: RTS – A Git-native execution provenance protocol for AI decisions
RTS: Git-native execution provenance protocol for documenting AI decision-making and reproducibility in AI systems.
3/1/2026
AI is making junior devs useless
Opinion piece questioning impact of LLMs on junior developer skill development and career progression.
3/1/2026
Show HN: Zagora, Distributed fine-tuning platform on mixed GPUs over internet
Zagora: distributed fine-tuning platform using pipeline parallelism for heterogeneous GPU clusters over standard internet.
3/1/2026
Show HN: Geostorm.ai – Monitor what AI chatbots say about your software
Tool monitoring AI chatbot recommendations across LLM models like ChatGPT and Claude. Tracks where products are mentioned by different AI systems.
3/1/2026
Simple Made Inevitable: The Economics of Language Choice in the LLM Era
Technical analysis of language choice tradeoffs in microservice architectures during LLM era, featuring Clojure and composition patterns.
3/1/2026
Show HN: I'm building a platform to manage larger projects with AI agents
Frame: terminal-first IDE platform for developing and managing larger AI agent projects with multi-AI support.
3/1/2026
Show HN: The framework for AI-native MCP servers
MCP Fusion framework for building AI-native servers with type-safe tools, presenters, and zero boilerplate. Includes scaffolding and Cursor integration.
3/1/2026
There's Only One Idea in AI
Historical perspective on neural network breakthroughs, discussing deep learning fundamentals like ReLU and gradient descent as core ideas in AI.
3/1/2026
Show HN: Auto-cleanup for Claude Code's orphan process memory leak
Automated cleanup tool for orphan Claude Code subagent processes and MCP servers that leak memory after session termination.
3/1/2026
I used 2D Base64 to bypass Gemini and expose Google's moderation flaws
Security research on bypassing Gemini's safety filters using Base64 encoding. Demonstrates moderation system flaws.
3/1/2026
Show HN: Rulegen – Auto-generate Claude.md and .cursorrules from your codebase
Tool auto-generating CLAUDE.md and .cursorrules files from codebase analysis without requiring LLM for AI coding assistant context.
3/1/2026
Show HN: Steward – a background agent that closes 80% low-risk noise
AI agent that auto-executes low-risk tasks by connecting GitHub/email/Slack/calendar signals with safety gates and natural-language briefings.
3/1/2026
Show HN: Vote-MCP" – a bit like Google Forms, but for AIs
Vote-MCP: REST API coordination primitive for autonomous AI agents to create polls and make collective decisions. Demo available.
3/1/2026
GPT detectors are biased against non-native English writers (2023)
Research paper showing GPT detectors have bias against non-native English writers, affecting fairness of AI detection systems.
3/1/2026
Show HN: CANomaly-LSTM – Detecting CAN bus anomalies with deep learning
LSTM autoencoder system for detecting anomalies in CAN bus traffic using deep learning for vehicle cybersecurity.
3/1/2026
Show HN: RAG-Enterprise – 100% local RAG system for enterprise documents
Open-source 100% local RAG system for enterprise document processing with one-command setup and data privacy guarantees.
3/1/2026
Give AI agents a real browser, watch them live via WebRTC
vscreen: WebRTC bridge enabling AI agents to control real browsers, click/type on websites, with 63 MCP automation tools and live streaming.
3/1/2026
Show HN: Aide – Opinionated, deterministic code editing for AI agents
AIDE: CLI tool for AI agents to perform deterministic code analysis, refactoring, and generation with automated test verification and rollback.
3/1/2026
Show HN: AgentLens – Open-source observability for AI agents
Open-source observability platform for AI agents with topology visualization, time-travel replay, and multi-agent debugging. Self-hosted alternative to LangSmith.
3/1/2026
Hackerbot-Claw: AI Bot Exploiting GitHub Actions – Microsoft, Datadog Hit So Far
Autonomous AI bot conducted week-long attack on GitHub Actions CI/CD pipelines, achieving RCE and token exfiltration across major repositories.
3/1/2026
Intelligence is a commodity. Context is the real AI Moat
Summary from AI meetup discussing intelligence commoditization, context as competitive advantage, and agent security.
3/1/2026
Show HN: Optimal: Cost effective infra with agentic inbox
Platform using AI agents to generate cost-optimal infrastructure plans for ML workloads with agentic inbox for user interaction.
3/1/2026
User Privacy: An Analysis of Frontier LLM Privacy Policies (2025)
arXiv research analyzing privacy policies of frontier LLMs (2025). Evaluates user data handling across major models.
3/1/2026
Show HN: OneCamp – Self-Hosted Slack/Asana/Zoom/Notion Alternative
Open-source Slack/Asana/Zoom/Notion alternative with real-time CRDTs, WebRTC, LiveKit SFU, and Python AI agent for audio transcription.
3/1/2026
Show HN: Ghostty Pane Splitter – Split terminal panes for AI coding agents
Rust CLI tool for creating multi-pane terminal layouts optimized for running multiple AI coding agents in parallel.
3/1/2026
Update: Making VeriContext Enforce Citations Across Sub-Agents
Update on VeriContext enforcing citations across sub-agents. Limited content detail available.
3/1/2026
10-202: Introduction to Modern AI (CMU)
CMU course introducing modern AI systems, machine learning methods, and LLMs behind ChatGPT, Gemini, and Claude.
3/1/2026
Show HN: MemLineage: governed writes for AI agents
OpenClaw infrastructure providing governed memory and task management for AI agents with PR-like control loop for write safety.
3/1/2026
Show HN: FeatureDrop – Open-source client-side product adoption engine (3 kB)
Open-source lightweight product adoption toolkit with feature tours and feedback. React/Vue/Svelte compatible, 3KB core.
3/1/2026
Show HN: ExamJungle – AI-driven exam simulators for IT certifications
Educational platform for IT certification prep using AI to explain concepts and adjust difficulty. LLM integration for learning.
3/1/2026
Show HN: External Threat Protection in GitHub Agentic Workflow
GitHub Agentic Workflow feature allows writing CI/CD pipelines in English markdown instead of YAML, using AI agents with repository context.
3/1/2026
Roast My Code – AI that scores and roasts your codebase (open source)
Open-source AI tool using Groq/NVIDIA NIM to analyze and humorously critique code quality. LLM-powered code review.
3/1/2026
Show HN: Agentic Workflows – 56 Ready-to-use Templates
56 ready-to-use GitHub workflow templates for automation tasks like issue triage, PR review, and release notes generation.
3/1/2026
Show HN: Gpu.fund – GPU rental prices across Vast.ai, RunPod, AWS, GCP and more
GPU.fund is a comparison tool for GPU rental prices across cloud providers like Vast.ai, RunPod, AWS, and GCP, tracking H100, A100, and RTX 4090 pricing.
3/1/2026
Show HN: Papercut – track ArXiv topics, get notified, skim with AI summaries
Papercut is an app that tracks ArXiv papers in selected topics, sends notifications for new papers, and provides AI-generated summaries in multiple formats (TL;DR, math breakdown, ELI5, methodology).
3/1/2026
Show HN: I built an open-source D&D app using Python and Llama 3.1
Open-source Python web app using Llama 3.1 to automate D&D campaign management and reduce game master prep burden.
3/1/2026
Show HN: SkillFortify, Formal verification for AI agents (auto-discovers)
Formal verification tool for AI agent skills. Auto-discovers malicious tools across 22 agent frameworks with zero configuration.
3/1/2026
Show HN: I built a desktop app combining Claude, GPT, Gemini with local Ollama
Desktop application (PyQt6) that combines cloud LLMs (Claude/GPT/Gemini) with local Ollama models in a 3-phase pipeline: task decomposition, local processing, result integration.
3/1/2026
Show HN: Iosef, an iOS simulator CLI designed for agents
iOS simulator CLI tool designed for AI agents. Enables LLMs to control simulators with automatic screenshot scaling and selector-based interactions.