25d ago
Show HN: Forgeterm – Runtime security monitor for AI coding agents
Forgeterm is a runtime security monitor designed specifically for AI coding agents.
25d ago
Show HN: Scindo, AI agent that turns team discussions into shipped code
Scindo: AI agent converting team discussions into code PRs. Agentic workspace managing conversation-to-code workflow.
25d ago
Free Chrome extension to see what employees send to AI tools
Chrome extension monitoring employee data sharing with AI tools. Security/compliance tool for agent data protection.
25d ago
Show HN: OpenFable – Open-source RAG engine using tree-structured indexes
OpenFable: open-source RAG engine implementing FABLE algorithm with tree-structured indexes for retrieval across document hierarchies.
25d ago
LLM plays an 8-bit Commander X16 game using structured "smart senses"
LLM connected to 8-bit game via structured text summaries; maintains notes, develops strategies, discovers AI exploits.
25d ago
Show HN: Yolt – safer YOLO mode for LLMs (recover deleted or overwritten files)
Title only: Yolt tool for recovering deleted/overwritten files in safer LLM YOLO mode.
25d ago
Show HN: Heron is open-source security auditor that interviews your AI agents
Open-source MIT-licensed tool that audits AI agents through interview-style questioning with regulatory compliance flags (SOC2, GDPR, EU AI Act).
25d ago
Show HN: Flowcost – know what your AI workflow is likely to cost
Developer tool for estimating AI workflow costs accounting for retrieval, retries, tool use, and model choices.
25d ago
LLM inference engine from scratch in C++ – why output tokens cost 5x
Deep dive into LLM inference engine implementation in C++ explaining why output token generation costs 5x more than input processing.
25d ago
Show HN: Heron is open-source security auditor that interviews your AI agents
Open-source MIT-licensed tool that audits AI agents through interview-style questioning with regulatory compliance flags (SOC2, GDPR, EU AI Act).
25d ago
MegaTrain: Full Precision Training of 100B+ Parameter LLMs on a Single GPU
Research on full-precision training technique for 100B+ parameter LLMs on single GPU.
25d ago
The Empty Middle of AI Coding
Opinion piece on AI coding tools. Minimal technical detail provided.
25d ago
Hacking Google Support: Leaking call logs and deanonymising agents
Security vulnerability disclosure about Google Support chat widget leaking call logs and agent data.
25d ago
Anthropic's most powerful AI model Mythos Preview is too dangerous for release
News article reporting Anthropic's Mythos model withheld from release due to vulnerability detection capabilities.
25d ago
Writing an LLM from scratch, part 32i – Interventions: what is in the noise?
Training 163M-parameter GPT-2-style model from scratch; analysis of loss differences vs original GPT-2.
25d ago
Tell HN: I think my colleagues are just piping my replies into an LLM
Discussion post about colleagues using LLM-generated responses. Opinion piece without technical content.
25d ago
Pg_sage – Agentic Postgres Dba
Agentic PostgreSQL DBA tool running as Go binary alongside PostgreSQL, executing diagnostic rules and optional LLM-powered fixes with trust-ramped automation.
25d ago
Why LLMs Are Bad Writers but Good Editors
Essay contrasting LLM capabilities in technical domains versus creative writing, citing Altman and Cowen perspectives.
25d ago
Show HN: Voxcode: local speech to text and ripgrep = transcript and code context
Developer tool combining local speech-to-text with ripgrep code search, enabling voice-based code navigation for AI coding assistants.
25d ago
Breakthrough in AI Solving Math Conjectures: Peking Univ. Team's Exploration
Peking University AI framework autonomously discovered and formally verified solution to open problem in commutative algebra using 19,000 lines of Lean 4 code.
25d ago
Did it get dumber? Tracking Claude Code and Codex according to HN comments
Hacker News sentiment tracker for Claude Code and Codex coding tools, updated daily with community opinions on performance.
25d ago
OCC – YAML-defined workflow orchestrator for LLMs with parallel execution
Multi-model workflow orchestrator with YAML-defined chains, parallel execution, visual canvas builder, and MCP/REST/CLI interfaces supporting tool-based agents.
25d ago
Hugging Face Contributes Safetensors to PyTorch Foundation
Safetensors joins PyTorch Foundation as hosted project to secure model distribution and prevent arbitrary code execution in agentic solutions.
25d ago
Defender – Local prompt injection detection for AI agents (no API calls)
Local prompt injection detection tool for AI agents using MCP and function calling, runs ONNX model offline without API calls.
25d ago
Claude Mythos Preview: Available in Private Preview on Vertex AI
Announcement of Claude Mythos Preview availability on Google Cloud Vertex AI for select customers as part of Project Glasswing.
25d ago
Optimize Norm Gathering Asm
Technical post on optimizing norm gathering assembly code for SereneDB search benchmark across x86_64 and ARM architectures.
25d ago
ZeroKeep – Private AI workspace on WebGPU. Zero cost, zero setup
Browser-based private AI workspace running models via WebGPU with zero server communication, local storage, and no API key requirements.
25d ago
Show HN: DryRun – PostgreSQL MCP and schema advisor that works offline
PostgreSQL MCP server providing schema awareness to AI agents and coding assistants from offline snapshots without production database credentials.
25d ago
Mamba-3 and the State Space Model Renaissance
Research note on Mamba-3, a state space model architecture improving on transformers' quadratic attention costs for efficient LLM inference in agentic applications.
25d ago
Show HN: I built an open protocol for Agent-to-agent commercial negotiation
A2CN: open protocol for safe agent-to-agent commercial negotiation enabling procurement agents to transact deals.
25d ago
An AI Agent. In COBOL. Yes
COBOL-based AI agent chatbot with tool use and agentic loop, integrating with modern LLMs via OpenRouter.
25d ago
The Signature Method in Machine Learning (an interactive reading note)
Reading notes on signature method for feature engineering from sequential event data in machine learning.
25d ago
Completely Private AI
Self-hosted AI assistant with voice, vision, RAG, and web search running entirely on-device without cloud.
25d ago
FormTo – Open-source, self-hosted form back end
Open-source, self-hosted form backend alternative to SaaS solutions like Formspree with email/webhook delivery.
25d ago
HappyHorse 1.0 – Turn Any Idea into AI Videos Instantly
AI video generation tool converting text and images to videos for content creation platforms.
25d ago
What "Open" Means: Abliterating Gemma 4 in 24 Minutes
Automated decensoring tool reduced Google's Gemma 4 refusal rate from 98% to 47% in 24 minutes on laptop.
25d ago
Lunar Gateway or Moon Direct? (2019)
2019 op-ed critiquing NASA's proposed lunar Gateway space station design.
25d ago
Show HN: Rootcx.com – open-source AI agents and internal software
Open-source infrastructure for building AI agents and internal software with managed database, auth, and RBAC.
25d ago
Show HN: SharpSkill – We built the future of AI coding interviews
SharpSkill is an interview preparation platform for coding assessments covering React, Node.js, and other tech stacks.
25d ago
AI-Ready Modular Data Center Slashes Deployment Time
Modular data center pods offer faster deployment as alternative to hyperscale projects, relevant for AI infrastructure.
25d ago
Is Entire.io hype or is it the future of GitHub?
Analysis of Entire.io startup and its Checkpoints open-source CLI tool providing observability layer for AI coding workflows.
25d ago
Milla Jovovich released an AI memory system. None of benchmark scores are real
Investigation into false benchmark claims in MemPalace, an open-source AI memory project. Exposes fabricated scores and questions attribution to actress Milla Jovovich.
25d ago
Claude AI down: Anthropic users hit with errors as chatbot goes offline
Claude chatbot outage with elevated error rates affecting Sonnet 4.6 model and downstream services.
25d ago
Show HN: Bring AI Agents to industrial control via "SDK-style" real-time engine
Python real-time engine with sub-1ms jitter for industrial control, auto-generates REST APIs and MCP for LLM agent integration.
25d ago
Anthropic's latest AI model could let hackers carry out attacks faster
Anthropic provides Mythos model to major tech companies for cybersecurity testing and vulnerability discovery.
25d ago
Open-source weekly live coding session (Thur 5PM)
Open-source framework for AI SRE agents that integrate 40+ infrastructure tools to autonomously investigate and resolve production incidents.
25d ago
Why GitOps still matters in a world of AI agents (FluxCon) [video]
Video presentation on GitOps relevance and practices in systems managed by AI agents, from FluxCon conference.
25d ago
Yu – Sandboxes your Claude Code/Codex with zero credential exposure
Yu is a sandboxing tool that isolates Claude Code and Codex execution to prevent credential exposure from compromised code or dependencies.
25d ago
GLM 5.1: Pelican Test
GLM-5.1 is a 754B parameter open-source LLM that demonstrates improved reasoning and multi-modal capabilities like unprompted SVG+CSS generation.
25d ago
Your parallel Agent limit
Analysis of cognitive load and limitations when managing multiple parallel AI agents, focusing on human-in-the-loop costs beyond throughput metrics.