HN vibecoder21 2d ago

I collected startup ideas. It changed how I think about ideas completely

Personal essay about collecting startup ideas rather than obsessing over single concepts.

HN leonidas1712 2d ago

A3: Kubernetes for autonomous AI agent fleets

A3: Kubernetes-based platform for autonomous AI agent fleets at SAP Labs Singapore. Handles code, research, slides, patents, audits with task distribution and execution planning.

HN georgestrakhov 2d ago

Neural Computer

Research essay on whether AI can function as a computer and self-optimize its own execution.

HN markfront 2d ago

Show HN: RelayFreeLLM — Free AI Gateway with Auto-Failover (Updates)

RelayFreeLLM: open-source gateway aggregating free LLM APIs (Gemini, Groq, Mistral, Cerebras, Ollama) with automatic failover and OpenAI-compatible endpoint.

HN aatel-license 2d ago

Code Democracy the Big Lie

Opinion piece on whether AI coding tools like Claude Code democratize software development.

HN spaccy05 2d ago

Show HN: Provepy – A Python decorator that proves your code using Lean and LLMs

Provepy: Python decorator using LLMs and Lean for formal code verification. Makes formal methods accessible via English claims.

HN datalater 2d ago

Crawl Code – because prompting was a text adventure

Crawl Code: Dungeon crawler game interface for Ollama LLM interactions, providing gamified prompting experience.

HN anitial 2d ago

Hindsight – A design spec for self-improving LLM agents

Hindsight: Design specification framework enabling LLM agents to learn from mistakes across sessions and internalize lessons into permanent behavior.

HN pdappollonio 2d ago

Show HN: Dux, a TUI Agent Multiplexer

Dux: TUI multiplexer for running multiple AI agents on same codebase via git worktrees, supporting Claude and other agent backends.

HN vinniefalco 2d ago

Brief: Bitcoin Core Governance Analysis

Analysis of Bitcoin Core governance arguing the project has stalled despite surface-level activity metrics.

HN haoliangyu 2d ago

Cmdgraph – Document any structural CLI for humans and agents

CLI documentation tool that recursively introspects help commands and exports structured data (JSON, Markdown, HTML) for human and AI agent consumption.

HN matt_d 2d ago

Optimization of 32-bit Unsigned Division by Constants on 64-bit Targets

arXiv framework announcement for developing features; incomplete content about division optimization research.

HN mercat 2d ago

Your next 10 hires won't be human

Multica: Open-source managed agents platform converting coding agents into autonomous teammates that handle task assignment, progress tracking, and issue resolution.

HN lablnet 2d ago

Show HN: HyperFlow – A self-improving agent framework built on LangGraph

HyperFlow is self-improving agent framework built on LangGraph. MetaAgent automatically optimizes TaskAgent performance through feedback loops.

HN kerng 2d ago

The AI-Assisted Breach of Mexico's Government Infrastructure [pdf]

PDF document about AI-assisted breach of Mexico's government infrastructure. Minimal content provided.

HN zach22 2d ago

Show HN: Lmscan – Detect AI text and fingerprint which LLM wrote it (zero deps)

Lmscan detects AI-generated text and identifies source LLM using statistical features. Open-source, offline, zero dependencies.

HN agiacalone 2d ago

Bring Back Buddy

Community response to removal of 'buddy' feature from Claude Code v2.1.97. No official changelog provided.

HN ValentineC 2d ago

Enforcing new limits and retiring Opus 4.6 Fast from Copilot Pro+

GitHub Copilot Pro+ enforcing usage limits and retiring Opus 4.6 Fast due to infrastructure strain from high concurrency patterns.

HN xlayn 2d ago

AMD GPU LLM Performance Testing

Performance benchmarking data for AMD GPUs running LLM inference. Tests actual hardware performance against theoretical specifications.

HN caihongxu 2d ago

Show HN: Palmier – Dispatch and schedule AI agents from your phone

Palmier app schedules and monitors AI agents from phone. Runs agents locally on user's machine without cloud dependency.

HN Meeeeeep 2d ago

We open sourced KubeezCut, a free browser based editor integrated with Kubeez

KubeezCut is client-side video editor using WebGPU and WebCodecs. Runs entirely in browser without backend or uploads.

HN iLemming 2d ago

Show HN: I forced Claude to play Tetris in Emacs

Developer stress-tests Claude with Emacs Tetris via custom elisp-eval MCP tool. Demonstrates LLM-driven REPL integration with persistent state across calls.

HN teleforce 2d ago

RAG 40x faster using binary quantization (2024)

Brief reference to binary quantization technique for faster RAG systems. Lacks technical details or implementation specifics.

HN festafin 2d ago

Hormuz, an MCP-first forecasting engine for reproducible modeling

Hormuz MCP-first forecasting engine for hydrocarbon-nitrogen-water modeling. Reproducible research stack with public MCP endpoint.

HN knivets 2d ago

Anthropic Mythos 0-days replicated with GPT5.4

Brief reference to Anthropic security vulnerabilities replicated in GPT5.4. Insufficient detail provided.

HN askalf 2d ago

Dario – use your Claude Max subscription as an API (no API key needed)

Dario tool converts Claude subscription into local API endpoint compatible with multiple frameworks. Supports all Claude models with native billing.

HN visionscaper 2d ago

Show HN: Collabmem – the memory system for long-term collaboration with AI

Open-source memory system for persistent human-AI collaboration over extended periods. Simple installation via MCP for long-term Claude interactions.

HN dividedcomet 2d ago

Open Source LLM Comparison – Is Opus Cooked?

Benchmark comparison of open-weight LLM models tested on identical prompts with cost and capability metrics. Tests latest frontier models on real-world tasks.

HN samchon-github 2d ago

Show HN: Function Calling Harness: From 6.75% to 100% (Qwen Meetup Presentation)

Function calling success rates for LLMs improved from 6.75% to 100% using structured output techniques. References EMNLP 2025 and ICLR 2025 benchmarks on nested tool calls and constrained decoding.

HN pabs3 2d ago

AI Code Is Hollowing Out Open Source, and Maintainers Are Looking the Other Way

Analysis of LLM-generated code integrated into open source projects and copyright/licensing implications for project sustainability.

HN thunderbong 2d ago

Daily You: A Personal Diary

Personal diary app with mood tracking and journaling features. Open-source but unrelated to AI/ML interests.

HN zeulewan 2d ago

Use Claude Max subscription with OpenClaw again

Tool to integrate Claude Max subscription with OpenClaw framework, bypassing Anthropic detection triggers.

HN tank-34 2d ago

Yet another BFF toolkit for Python

Python library for composing nested APIs declaratively with auto-batching, DataLoader pattern, and GraphQL generation.

HN Wuzu 2d ago

AI Agent Proxy to help reduce token usage (Anthropic Only)

Proxy system for Anthropic Claude that reduces token usage for AI agents.

HN bgrainger 2d ago

AI-Assisted Breach of Mexico's Government Infrastructure Report

Technical report on AI-assisted breach of Mexican government infrastructure resulting in exfiltration of citizen records.

HN moodiverse 2d ago

Most startup launch platforms are built backwards

Platform for posting startup projects without approval process or launch timing requirements.

HN simjnd 2d ago

Maki – the efficient coder (AI agent)

Rust-based AI coding agent with context token reduction techniques achieving 40% cost reduction and 2x speedup via skeleton parsing.

HN iljanevo 2d ago

Hit Claude usage limits? Resume your AI coding session elsewhere

Tool for resuming Claude AI coding sessions across rate limit boundaries.

HN wewewedxfgdf 2d ago

Faiss a library for efficient similarity search and clustering of dense vectors

Faiss library for efficient similarity search and dense vector clustering in C++/Python/GPU, developed at Meta AI Research for billion-scale vector retrieval.

HN atripat6 2d ago

Show HN: Ark – AI agent runtime in Go that tracks cost per decision step

Go-based AI agent runtime (ARK) with dynamic context optimization, adaptive execution, and cost attribution per decision step.

HN JunePanda 2d ago

Anthropic injects reasoning_effort=25 into Claude.ai consumer system prompts

Anthropic adds reasoning_effort parameter to Claude.ai consumer system prompts.

HN YutingZhong 2d ago

Show HN: TopRank – Open-Source Claude Code Skills for SEO and Google Ads

Open source Claude Code skills providing AI agents direct access to Google Search Console and Ads for SEO optimization and ad spend analysis.

HN asparagui 2d ago

Image Recognition with Lean 4

Using Lean 4 as specification language for neural networks with StableHLO/MLIR compilation to GPU via IREE, computing gradients at codegen time without Python runtime.

HN Rial_Labs 2d ago

Cisco got hacked through a security scanner

Cisco breach in 2026 using credentials from Trivy supply chain compromise, exposing source code for AI products across 300+ GitHub repositories.

HN tomflitt 2d ago

Show HN: A free study guide for the AWS DVA-C02, built from my own exam notes

Free study guide for AWS DVA-C02 certification exam created from personal notes using Claude for content formatting.

HN earayu 2d ago

A public Agent Sandbox with Hermes inside

Public sandbox environment for testing AI agents using Hermes model.

HN burgerguyg 2d ago

Linux/Mac Alternative to Nvidia Broadcast?

Open source GPU-accelerated Linux/Mac alternative to NVIDIA Broadcast providing background blur, virtual backgrounds, and noise cancellation.

HN keniz 2d ago

I built an AI that teaches your slides and answers questions

Lectura: AI tool that converts slides into reusable interactive presentations with language support and Q&A capabilities.

HN beacharsiu 2d ago

I Tried Giving an AI Agent Gmail Access. It Took 19 Steps and Still Failed

Technical analysis of limitations when giving AI agents Gmail access: OAuth, 2FA, browser automation, and privacy concerns in practice.

HN sethbannon 2d ago

What's Going on in AI

Elicit CEO discusses AI R&D progress, predicting AI researcher parity around 2030. Investor update excerpts on scaling AI companies.