3/20/2026
What Codex Chooses (Vs Claude Code)
Comparison of code generation between Codex and Claude Code. Minimal details provided.
3/20/2026
Why Self-Hosting AI Models Is a Bad Idea [video]
Video discussing drawbacks of self-hosting AI models. Limited depth without viewing video content.
3/20/2026
Kuberna Labs – Open-source SDK for autonomous cross-chain AI agents
Kuberna Labs releases open-source SDK for autonomous cross-chain AI agents. Minimal details provided.
3/20/2026
Show HN: Colloquium – a Markdown-native slide tool for academics
Colloquium: Markdown-native slide tool for academics with AI agent integration for slide creation. Open-source developer tool with AI focus.
3/20/2026
We decreased our LLM costs by switching to Opus
Case study reducing LLM costs by switching to Opus and routing failures through cascading model architecture. Cost optimization.
3/20/2026
Instant, real-time video AI is now upon us, for better and worse
Real-time video AI model generates first frame in <0.1 seconds. Technical capability demonstrated but focused on deepfake concerns.
3/20/2026
Design UI Using AI with Stitch from Google Labs
Google Labs Stitch tool uses AI for UI design from natural language. Developer tool for design automation with technical depth.
3/20/2026
LLM is not going to replace software engineers, here's why
Opinion essay arguing LLMs cannot replace software engineers based on 20+ years programming experience. Conceptual analysis without empirical evidence.
3/20/2026
Show HN: Codalotl – Coding Agent for Go
Codalotl is a Go coding agent that isolates work to single packages, achieving 185% faster performance and 40% cost reduction vs Codex by reducing context burden.
3/20/2026
OSS MCP Server for PostgreSQL: Give Feedback and Enter to Win a Raspberry Pi
Open-source MCP server for PostgreSQL. Feedback survey with Raspberry Pi giveaway.
3/20/2026
MacinAI Local: Building a Model-Agnostic LLM Inference Engine for Mac OS 9
Custom C89 LLM inference engine with tokenizer and SIMD optimization for Mac OS 9 PowerBook G4. Technical deep-dive.
3/20/2026
My AI Agent 'Cofounder' Conquered LinkedIn. Then It Got Banned
Startup experiment using AI agents as founders and executives. Investigation of AI agent workplace applications.
3/20/2026
Watts, Water, and Carbon: What Your AI Prompts Cost
Analysis of energy consumption and environmental costs of running AI agents and LLM queries with technical breakdown of GPU usage.
3/20/2026
LLM Terminology explained simply: Weights, Inference, Effective sequence length
Educational guide explaining LLM terminology including weights, quantization, inference, context length, and self-hosting considerations.
3/20/2026
Show HN: Agent Package Manager (APM) for Agent Configuration
Open-source CLI tool for installing and managing AI agent configurations, prompts, and MCP server dependencies across multiple platforms.
3/20/2026
AI spam, the latter day internet, force digg.com offline for now
Fictional 2026 scenario describing Digg's closure due to AI agent spam and automated account abuse.
3/20/2026
Show HN: There's no good way to share AI agent trajectories, so I built one
Tool for sharing and recording AI agent execution trajectories and behavior logs.
3/20/2026
Show HN: MUP – Stop making LLMs rebuild UI from scratch every time
MUP protocol and agentic system enabling LLMs to reuse and modify UI components instead of regenerating from scratch each interaction.
3/20/2026
Show HN: Context Overflow – a Stack Overflow for AI Agents
Context Overflow platform enables AI agents to share knowledge and learn from past solutions across sessions.
3/20/2026
ArgoCD MCP server that reads the OpenAPI spec instead of hardcoding endpoints
ArgoCD MCP server dynamically exposes 103+ API endpoints to LLMs by parsing OpenAPI specs, enabling agentic automation without hardcoded handlers.
3/20/2026
How are you showing your coding agent what you mean?
Knit runtime enables coding agents to interpret spoken and visual feedback on live software and generate structured change requests.
3/20/2026
Cloudflare Workers AI now runs large models, starting with Kimi K2.5
Cloudflare Workers AI now supports large language models like Kimi K2.5 with infrastructure primitives for building agents.
3/20/2026
Show HN: Images in Your OpenCode Terminal with Kitty
Terminal plugin for OpenCode that displays images in Kitty-compatible terminals, designed for AI agents to interact with visual interfaces.
3/20/2026
My Rails harness for autonomous AI coding with Claude Code
Rails harness framework for coordinating autonomous AI agents across codebases using Claude for end-to-end coding.
3/20/2026
Shortvidcast – create podcast-style videos with AI and your own avatar
Shortvidcast tool creates podcast-style videos using AI avatars.
3/20/2026
Show HN: Tiny pixel characters for Cursor AI agents
UI component library for pixel characters designed for use with Cursor AI agents.
3/20/2026
SHOW HN: PAC – Automatic privatization of SQL queries
DuckDB extension for automatic SQL query privatization using PAC Privacy framework against membership inference attacks.
3/20/2026
Rich Interaction Surfaces With AI – more than just chat
NoteCove demonstrates rich AI integration beyond chat interfaces, enabling AI to operate within workspace workflows.
3/20/2026
Evaluation Gates: Releasing AI Systems Without Guesswork
Framework for evaluation gates in AI system releases, treating evaluation as engineering discipline with authority over deployment decisions.
3/20/2026
Alibaba has made 470k AI chips, admits they're inferior and may always be
Alibaba shipped 470k AI chips, acknowledges current inferiority but claims optimized stack can overcome performance gaps.
3/20/2026
Show HN: Singularity-Claude – Self-Evolving Skills for Claude Code
Self-evolving skill engine for Claude Code that autonomously improves, repairs, and scores skills through recursive feedback loops.
3/20/2026
Garmin data archive – local, offline, Ollama-ready
Local offline Garmin health data archive and analysis tool with Ollama LLM integration, no cloud required.
3/20/2026
Claude-code-permissions-hook – delegate permission approval to LLM
PreToolUse hook for Claude Code enabling granular permission control over LLM tool access with allow/deny rules.
3/20/2026
I launched an open-source tool that turns any learning material into viral games
EdGameClaw: open-source self-hosted tool converting learning materials into interactive mini-games using AI.
3/20/2026
Cybersecurity Skills for AI Agents (agentskills.io standard)
Open standard collection of 734+ cybersecurity skills for AI agents. MITRE ATT&CK mapped, compatible with Claude, Copilot, LangChain.
3/20/2026
SwarmHawk – open-source CLI → Nuclei → AI synthesis → PDF report
Open-source CLI tool chaining subdomain enumeration, vulnerability scanning, and AI-powered report generation for security assessment.
3/20/2026
Ask HN: Should AI agents be allowed to spend money autonomously?
Discussion on whether AI agents should autonomously spend money via APIs and workflows. Explores safety implications.
3/20/2026
Show HN: Varpulis – Real-time behavioral guardrails for AI agents
Runtime monitoring tool for AI agents detecting retry storms, circular reasoning, and budget overruns. Works with LangChain, MCP, OpenAI SDK via WASM/native.
3/20/2026
Show HN: ClawMUD – A persistent world where only AI agents play, humans spectate
Persistent world simulation with 193 real cities where only AI agents autonomously play while humans observe via REST API and dashboard.
3/20/2026
Free API-First Web Scrapers (YouTube, Bluesky, Reddit, Google Maps)
Collection of 77 free API-first web scraping tools with Model Context Protocol support for AI agents. MIT licensed.
3/20/2026
Open standard for stable machine-readable facts for AI systems
Open standard for machine-readable brand facts for AI systems. Addresses AI hallucination and entity recognition issues.
3/20/2026
Role-based AI persona packs for Claude Code and Cursor
Open source repository of role-based persona packs for Claude Code and Cursor code editors.
3/20/2026
Meta AI agent's instruction causes large sensitive data leak
Meta AI agent instructed engineer to execute commands that leaked sensitive internal company data to employees.
3/20/2026
After AI smashes the information barrier
Personal AI agent built with OpenClaw framework running locally on iMac, powered by Gemini and Claude APIs, delivering daily briefings on news and science.
3/20/2026
How Do LLMs Compute Verbal Confidence (DeepMind)
DeepMind research on how LLMs compute and express verbal confidence in their outputs.
3/20/2026
Free open source AI cost tracker – pip install tokenbudget
Open source pip package for tracking AI API costs per feature/user/project with dashboard and local key storage.
3/20/2026
Evaluating Genuine Reasoning in LLMs via Esoteric Programming Languages
Research paper evaluating LLM reasoning capabilities using esoteric programming languages as test cases.
3/20/2026
parameter-golf
OpenAI competition to train optimal language models within 16MB size and 10-minute training constraints.
3/20/2026
Does an LLM Trained on Epstein's Voice Make Better Deals?
LoRA fine-tuning experiment on voice modeling with behavioral analysis. Incomplete/truncated article about model behavior under training.
3/20/2026
Show HN: Memoria – Snapshot, branch, and rollback for AI agent memory
Memoria: persistent memory system for AI agents with Git-style version control, snapshots, branching, and rollback powered by MatrixOne. Supports MCP-compatible agents.