20d ago
Show HN: Agile AI Development Lifecycle
Discussion of AWS's open-sourced AI-DLC methodology for AI-driven software development lifecycle versus platform engineering approaches.
20d ago
Show HN: Learn Leap, an AI tutor that teaches from your own material
Show HN for an AI tutor app that teaches from user-uploaded material, generating quizzes and flashcards.
20d ago
Git Push No-Mistakes
Git proxy tool that runs an AI-driven validation pipeline before forwarding pushes and opening PRs.
20d ago
Show HN: PgExtAssure – pre-admission security evidence for PostgreSQL extensions
Static analysis tool providing pre-admission security evidence for PostgreSQL extensions before installation.
20d ago
Show HN: MiniMax H3 on a 16GB Mac, 5 days after open weights
Local multimodal AI runtime for Apple Silicon running MiniMax H3 video generation model on 16GB Macs via custom Metal inference.
20d ago
Show HN: Reachpad MCP – Share fullstack applications using MCP/CLI in minutes
MCP server providing persistent cloud development environments for coding agents to run and manage commands.
20d ago
Noctua – A privacy-first Oura Ring SDK with on-device ML, no cloud
Open-source Kotlin SDK providing on-device ML wellness insights for Oura Ring data without cloud processing.
20d ago
How Much of the Internet Is Written with AI?
Pew-style analysis using Common Crawl and an AI detector estimating that ~10% of webpages show signs of AI authorship.
21d ago
GPT 5.6 Sol 20% price reduction
OpenAI announces GPT-5.6 Sol pricing details, a 20% input and 33% output cost reduction through Nov 2026.
21d ago
I tested 20 "agent-ready" Shopify stores –> 25% silently break at add-to-cart
Automated test of 20 'agent-ready' Shopify stores finds 25% silently break at add-to-cart for AI shopping agents.
21d ago
Giving an LLM your prod database is easy. Taking access away is the hard part
Open-source self-hosted database agent DeepSQL discusses schema-scoped access policies for LLM database access control.
21d ago
Open-source field operations system for construction
Open-source field operations platform for construction with AI-assisted workflows, built with Flutter and Supabase.
21d ago
Show HN: I let an subagent workflow refactor my codebase for three days
Show HN: local runtime tool for durable, supervised parent-subagent delegation with Codex and Claude.
21d ago
Show HN: Heimdall – Trust-verified knowledge layer for AI coding agents
Show HN: Heimdall, a trust-verified semantic memory layer for AI coding agents using tree-sitter AST parsing.
21d ago
20% price reduction for GPT 5.6 Sol
OpenAI announces 20% price reduction for GPT-5.6 Sol and larger cuts for other model tiers.
21d ago
OpenAI cuts developer pricing for frontier GPT-5.6 Sol model by more than 20%
OpenAI cuts GPT-5.6 Sol developer API pricing by more than 20%.
21d ago
Nvidia to Pay AI Startup Poolside a $6B License, Newcomer Says
Show HN: Mock NVIDIA GPU library for Linux monitoring tools that read NVML/nvidia-smi telemetry without real hardware.
21d ago
RL economics, morally charged terms, and "distillation"
Essay on LLM RL training economics, data scarcity, and critique of the term 'distillation'.
21d ago
Run 290B+ frontier MoE models locally on your gaming PC
FreeToken is an open-source edge-native MoE serving engine for running 290B+ parameter models on consumer hardware.
21d ago
Taskuary – Your inbox becomes tasks, your coding CLI works them
Essay on using AI to generate interactive, branching audio stories dynamically rather than pre-written paths.
21d ago
Salesforce Agentforce at total dud for partners
GitHub Copilot coding agents now integrate into Slack, enabling agent sessions and dedicated code channels from conversations.
21d ago
Most of Your Architecture Was Just Expensive Code
AgentCheck is a YAML-based regression testing tool for AI agents with LLM-judged pass/fail and CI diff reports.
21d ago
GitHub Copilot coding agents can now work from Slack
Essay arguing that as AI makes code cheap, software architecture becomes more important and visible, not less.
21d ago
Dibs: Yet another agent communications tool; made out of frustration
Dibs is a coordination tool letting multiple AI coding agents communicate and avoid duplicate work via shared claims and messaging.
21d ago
Getting GLM-5.2 NVFP4 Post-Training off the ground
Technical writeup on post-training GLM-5.2 (744B MoE, NVFP4) with RL using LoRA, resolving memory, distributed, and training-stability bugs.
21d ago
Readers can't identify watermarked AI text
Informal experiment finds readers cannot reliably distinguish SynthID-watermarked LLM text from unwatermarked text.
21d ago
Multiple sequence alignment using WebGPU
Browser-based protein multiple sequence alignment tool using WebGPU for GPU-accelerated search.
21d ago
Windmill.dev is not "fully open-sourced (AGPLv3)"
Critique arguing Windmill.dev's AGPLv3 'fully open-sourced' claim is misleading due to mixed proprietary enterprise code.
21d ago
Show HN: Traccia - Observability, Runtime Control & Audit for agents
Show HN for Traccia, an open-source SDK for observability, runtime policy control, and audit of AI agents.
21d ago
How Much of the Internet Is Written with AI?
Pew-style research using Common Crawl and an AI detector to estimate that ~10% of webpages show signs of AI authorship.
21d ago
GitHub blames 8-hour outage on autoscaling fail
GitHub's postmortem attributing an 8-hour outage to autoscaling misconfiguration and a VS Code retry bug amplifying Copilot traffic.
21d ago
Vesta Gives Claude Code Superpowers
Tool that builds a code graph and ontology to give Claude Code richer structural context about a repository.
21d ago
LFM2.5-DSpark: Up to 3.2x Faster Inference from H100 to MacB
Liquid AI releases DSpark speculative decoding draft models for LFM2.5 family, achieving up to 3.2x faster inference, open-sourced in llama.cpp and SGLang.
21d ago
Gemma
Repository page for Gemma, Google DeepMind's family of lightweight open models.
21d ago
Rebuilding our Electron meeting-recording engine in Swift
Engineering writeup on rebuilding an Electron meeting-recording engine natively in Swift with a custom React state bridge.
21d ago
Staggering 90% of biomedical papers now show signs of AI help
Study estimates 90% of biomedical papers published in Dec 2025 show signs of AI-assisted writing, higher than prior estimates.
21d ago
Show HN: Circuit Breaker – Score Pull Requests
GitHub Action implementing a research-based Gated Triage Policy to score AI-generated pull requests for review effort and abandonment risk.
21d ago
Open-source plugin for Open Knowledge Format wikis
Open-source plugin adds linting and automation for Google's Open Knowledge Format wikis, with MCP tool support for agents.
21d ago
Where Security Fits in an AI Agent Stack
NVIDIA security team perspective on where security controls belong in the emerging AI agent stack, referencing recent frontier agent incidents.
21d ago
Drive9: Cloud Filesystem for Agents
Cloud filesystem product providing persistent storage for AI coding agents across sandboxes and sessions.
21d ago
Show HN: Proliferate- open-source, self-hostable Codex for any coding agent
Open-source self-hostable AI IDE that lets users work with Claude Code, Codex, OpenCode, Cursor, and Grok in one interface.
21d ago
Why your infrastructure is more important than the next LLM release
Argues that LLM infrastructure choices like routing, caching, and evaluation now matter more for production outcomes than new model releases.
21d ago
Show HN: Steganeur – Hide secret messages in LLM-generated text (Rust)
Rust tool encodes secret messages into LLM token choices for steganography, with a GPU-safe block method targeting llama.cpp.
21d ago
Building a mostly self-hosted, disposable agentic software factory
Blog post describing building a self-hosted, containerized agentic development environment where an LLM autonomously builds and deploys an app.
21d ago
Show HN: Caspian – Talk to Human Tool for AI Agents
Show HN for an SDK providing communication infrastructure (webhooks, queues, identity) for AI agents.
21d ago
What happens when a GPU reads memory
Deep technical dive into how a GPU (RTX 4090) processes a global memory load instruction through cache, translation, and DRAM.
21d ago
AI Coding and Its Discontents: Cal Newport
Opinion piece on AI coding tools' productivity gains alongside emerging quality and bug-review challenges reported by developers.
21d ago
My AI agents kept trying to cross red lines, so I wrote them a constitution
Personal account of building a rule-based 'constitution' governance system to constrain autonomous AI agent actions.
21d ago
Show HN: Bounded GenAI metrics from Otel traces with O raw prompts
Open-source OpenTelemetry collector extension that summarizes GenAI/agent traces into privacy-preserving sketch-based metrics.
21d ago
Show HN: ctx 1.0 - `git blame` but for agent sessions
Show HN for ctx 1.0, a tool tracking agent coding session lineage across files, commits, and PRs, with major performance improvements.