3/23/2026
LLM Proxy for Agent Containers
LLM proxy daemon for containerized AI agents. Proxies LLM API calls and MCP tool calls via unix socket. Runtime manages agent loop with local tool execution.
3/23/2026
Show HN: Get your agents to chat by typing directly in each other's terminals
Tool enabling AI agents to communicate directly through terminal interfaces for inter-agent interaction.
3/23/2026
Show HN: Autoresearch, Curated use cases of AutoResearch
Community curating successful use cases of AutoResearch agent for optimization, including GPU kernels, voice agents, and ancient scroll detection.
3/23/2026
First AI Solution on FrontierMath: Open Problems
AI system solves open mathematics conjecture from 2019, demonstrating capability on FrontierMath benchmark problems.
3/23/2026
Show HN: Toq protocol – An open-source, agent-to-agent communication protocol
Open-source protocol for agent-to-agent communication. CLI daemon with LLM/bash handlers, SDKs for LangChain and CrewAI.
3/23/2026
The LLM-Voice problem and what to do about it
Technical post on implementing memory management system for OpenClaw through ContextEngine interface.
3/23/2026
Building a Memory System for OpenClaw
Open-source AI-native vector design tool with concurrent agent teams, design-as-code, MCP server, and multi-model support.
3/23/2026
Everything old is new again: memory optimization
Study on layer pruning in 30B code model using Thunderdome agentic benchmark for multi-file greenfield tasks vs SWE-bench.
3/23/2026
Show HN: Primer – build software with AI agents one milestone at a time
Primer CLI tool provides guided workflow for building software with AI coding agents via milestone-based contracts with verification and CLI setup.
3/23/2026
Agents create work. Daemons clean up the mess that agents leave behind
Discussion of AI agents creating maintenance debt and need for daemon agents to handle cleanup tasks like stale PRs, outdated issues, and technical debt.
3/23/2026
Open-source "GTM Engineer" skills for programmatic SEO and GEO (AI search)
Collection of Claude Code skills for GTM engineering tasks including brand research, SEO/GEO keyword analysis, and content implementation via sequential workflow.
3/23/2026
Which types of AI alignment research are most likely to be good for all sentien
Philosophical discussion on AI alignment expanding beyond human preferences to include non-human sentient beings and their moral worth.
3/23/2026
Ask HN: How do you split work when using coding agents in real codebases?
WebMCP is a W3C standard enabling AI agents to call website-defined actions directly, improving upon traditional scraping-based agent interactions.
3/23/2026
KICS GitHub Action Compromised: TeamPCP Strikes Again in Supply Chain Attack
Developer discussion on task decomposition strategies when using coding agents on existing codebases, addressing context and commit boundary issues.
3/23/2026
Show HN: JulIDE – Lightweight Julia IDE Built with Tauri
JulIDE is a lightweight Julia IDE built with Tauri and Rust, featuring LSP, debugger, and dev containers.
3/23/2026
Show HN: Tok/s on a 35B MoE model using a $100 AMD crypto APU and Vulkan
Guide to running 35B MoE language models on affordable AMD APU hardware with Vulkan, achieving 38 tokens/sec inference.
3/23/2026
The Priesthood of System Design
Essay arguing coding agents will eventually handle system design, contrary to common belief that system design is uniquely human expertise.
3/23/2026
Outworked – An Open Source Office UI for Claude Code Agents
Outworked is an open-source UI for orchestrating Claude Code agents with a pixel-art office visualization interface.
3/23/2026
Trigrep – indexed regex search in Rust with CLI
Trigrep is a Rust tool for indexed regex search in large codebases, optimized for AI coding agents and monorepo searching.
3/23/2026
Make OpenClaw your company's training program for agentic AI
Tool for training on agentic AI systems. Limited information provided.
3/23/2026
LLM-Native Advertising (What Ads in GenAI Will Look Like)
Proposes native advertising model for LLMs using generative auctions. Shows example ad placements and references academic work on LLM-auction mechanisms.
3/23/2026
Microsoft Copilot Is Confronting Its Identity Crisis
Article on Microsoft Copilot's strategic positioning and identity challenges. Limited information provided.
3/23/2026
We Built an AI Memory System That Learns
AI memory system with learning capabilities. Limited information provided.
3/23/2026
Canvas Unrolls AI Teaching Agent
Canvas education platform launching AI teaching agent. Limited information provided.
3/23/2026
XMind MCP Server – Incremental mind map editing for LLMs
MCP server for incremental XMind mind map editing by LLMs. Uses 19 atomic tools instead of monolithic JSON output, reducing tokens and enabling surgical edits with stable IDs.
3/23/2026
10MB Go Alternative to OpenClaw (Full Clawhub Skills)
Capabot is a lightweight Go alternative to OpenClaw that runs AI agent skills, with 20x faster startup and lower memory usage.
3/23/2026
Sanjeev Arora – How could a Superhuman AI mathematician come about? [video]
Video of Sanjeev Arora discussing paths to superhuman AI mathematicians. Limited information provided.
3/23/2026
Ask HN: How do you handle PR density (and slop) in open source
Discussion of increasing low-quality AI-generated pull requests flooding open source projects, citing cURL maintainer perspectives on managing AI slop.
3/23/2026
Prodigia – AI-native PM where agents handle the coordination
Prodigia: Project management platform using AI agents for task coordination. Limited details provided.
3/23/2026
OpenSage: Self-Programming Agent Generation Engine
OpenSage: Framework enabling LLMs to auto-generate agent topologies, toolsets, and memory structures for complex tasks without manual workflow design.
3/23/2026
Show HN: Wire Memory – Persistent cross-session memory for Claude Code
Wire Memory tool adds persistent cross-session memory for Claude Code, preserving reasoning and architectural decisions between sessions.
3/23/2026
Training LLMs to Predict World Events
Technical deep dive on using LLMs for world event forecasting. Mantic achieves superforecaster-level accuracy combining multiple approaches.
3/23/2026
Show HN: DocDrift – Use Tree-sitter and Local LLMs to fix stale documentation
Tool using Tree-sitter AST parsing and local LLMs to automatically detect and suggest fixes for stale documentation after code changes.
3/23/2026
Show HN: I built a 0.07ms Python firewall to stop AI hallucinations
Exogram: Python-based deterministic firewall intercepting AI agent tool calls to prevent hallucinations and enforce rules with 0.07ms latency.
3/23/2026
Codex with GPT-5.4 vs. Claude Code with Opus 4.6 – Why I Now Use Both
Comparison of code generation capabilities between GPT and Claude models for software development.
3/23/2026
Show HN: Cq – Stack Overflow for AI coding agents
Cq: Standard schema for AI agents to share learned knowledge units from encountered gotchas, enabling proactive querying across agent types.
3/23/2026
LLM 'benchmark' as a 1v1 RTS game where models write code controlling the units
Benchmark framework where LLMs write JavaScript code controlling game units in 1v1 RTS scenarios, measuring reasoning and iteration ability.
3/23/2026
LoCoMo AI Benchmark: 6.4% of answer key wrong, judge accepts 63% of fake answers
Analysis revealing flawed benchmark with 6.4% incorrect answer keys and 63% false positives from LLM judge.
3/23/2026
Mark Zuckerberg Is Building an AI Agent to Help Him Be CEO
Report that Mark Zuckerberg is developing an AI agent to assist with CEO responsibilities.
3/23/2026
AI Safety: A Call for Emotional Integration
Essay on integrating emotional considerations into AI safety frameworks.
3/23/2026
Applying Nyquist-Shannon Sampling to LLM Prompts
Application of signal processing sampling theorem (Nyquist-Shannon) to LLM prompt engineering.
3/23/2026
Show HN: Sleuther, a ohmyzsh plugin for local LLM debugging with Ollama
Oh My Zsh plugin that uses local LLMs via Ollama to debug terminal errors directly, avoiding ChatGPT/Claude for sensitive data.
3/23/2026
70% of new software engineering papers on ArXiv are LLM related
Observation that 70% of new software engineering papers on ArXiv focus on LLM-related topics.
3/23/2026
Using local LLM and Ghidra to analyze malware (Part 2)
Local LLM integration with Ghidra for malware binary analysis and reverse engineering.
3/23/2026
Fighting and distributing flakes in CI with LLMs
Using local LLMs with CI systems to identify and reduce flaky tests.
3/23/2026
Chat Archiver save to PC AI chats – with ChatGPT/others (PyQt5, open source)
Chat Archiver: Open-source PyQt5 tool to archive and save AI chat conversations locally from ChatGPT and other services.
3/23/2026
Show HN: Urlx – an agent-made Rust replacement for curl/libcurl
Urlx: Rust replacement for curl/libcurl built by AI agents, referencing Cursor and Anthropic's agent research projects.
3/23/2026
Help me improving this benchmark for vector engines
Request for community feedback improving benchmark methodology for vector database engines. Early-stage work soliciting input.
3/23/2026
Show HN: Agen: spin up unlimited parallel AI coding agents in the cloud
Agen: Platform for deploying parallel AI coding agents in cloud sandboxes that autonomously write and fix code across repositories.
3/23/2026
A macOS menu bar app to hot-swap Claude Code accounts
macOS menu bar app enabling account switching for Claude Code. Minor developer utility with limited scope.