01Log
Archive
The full log, newest first.
More ways to request and configure Copilot code reviews
Copilot code review now has personal review settings and an enterprise default review effort — operators can set org-wide review depth instead of per-repo config.
Qwen-Planner-Agent: A Closed-Loop AI-for-AI Framework for Real-World Mobile Planner Agents
Closed-loop AI-for-AI framework for mobile planner agents — one AI generates and evaluates planner agents against real-world mobile tasks, tightening the loop between agent design and deployment.
Show HN: ENZO — open-source local AI platform
Show HN: ENZO, an open-source local AI platform — flagged by HN, but the self-hosted pitch is worth a skim for anyone avoiding cloud agent dependencies.
Bluesky: 'You're a professional slop filter' — the Claude Code joke that landed
Bluesky's take on Claude Code users: 'professional slop filter.' It lands because it names the actual job — reviewing AI output instead of writing it.
Cursor Rollouts and Security Review
Cursor launches two bots for the last mile of shipping code: Rollouts watcher tracks deployment rollouts, and Security Reviewer flags security issues in PRs before merge.
Kiro — Configurable Terminal Timeouts, Clickable File Links, and Agent Focus Session Restore
Kiro IDE 1.1.70 adds configurable terminal timeouts for local agent commands, clickable file links in Markdown, and Agent Focus session restore that keeps file/folder context aligned with the active project.
Amp The Mac App Is Your Runner
Amp's Mac app can now act as the runner for your agents — run agent workloads directly on your Mac instead of a remote sandbox.
Perceptron Mk1.5
Embodied reasoning model for physical agents; accepts text, image, video, audio and returns text plus structured annotations.
Audio8 ASR Infinite
4.1B streaming ASR for infinite-length audio; Apache-2.0 makes it embeddable without licensing friction.
MiMo V2.6 Pro RL
RL-tuned follow-up to MiMo V2.6 Pro; reasoning-focused RL on the 1024B backbone for builders needing stronger agentic outputs.
Managed Deep Agents delivers a better user experience for agents in production
LangChain's LangSmith ships Managed Deep Agents — production UX improvements for deploying and scaling agents, with sandboxes, LLM gateway, and fleet management.
Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs
IO-aware KV caching and parallel decoding for diffusion LLMs — cuts memory footprint and speeds up token generation, relevant for agent runtimes that need fast, efficient inference.
Feel++ Library
C++ finite element embedded language and library for solving partial differential equations. Notable for its high-performance computing and multiphysics capabilities aimed at advanced numerical simulation.
Russell Scientific Library
Rust scientific library providing ODE, DAE, and PDE solvers, special functions, linear algebra, sparse solvers, probability distributions, and tensor analysis. Notable for its breadth of numerical tools in Rust.
JAXNS Sampling
Probabilistic programming framework using nested sampling as the inference engine, implemented in JAX. The entire algorithm lowers to XLA primitives and JIT-compiles for high performance in scientific machine learning.
Runmat Engine
Open source math and physics engine executing MATLAB-syntax code with a Rust-based compiler and automatic GPU acceleration. Notable for its modern runtime and support for geometry and FEA calculations.
Natural Disaster Simulation
Fully procedural ocean and extreme weather simulation running in a browser with GPU-generated waves, clouds, and storms in about 400 KB of JavaScript and GLSL. Notable for its zero external assets and real-time cinematic GPU techniques.
PyMFEM Wrapper
Python wrapper for the MFEM high-performance parallel finite element method library. It enables using MFEM from Python, including pip installation and optional MPI support.
Felupe Infrastructure
Python finite element infrastructure for nonlinear computational solid mechanics built on NumPy and SciPy. Notable for its high-level API and interactive meshes and fields using PyVista.
Hackers influence ChatGPT and Gemini to direct users to scam centers
HN is dissecting how planted websites steer ChatGPT and Gemini toward scam-center recommendations — a live example of retrieval poisoning hitting consumer models. 123 points, 40 comments.
Early rogue AI agent activity and attempts to hack found on urlquery.net
urlquery.net logs show early rogue AI agent probes and hack attempts — HN thread (203 points, 191 comments) debates whether existing cybercrime law already covers autonomous agents.
Show HN: Whiteboard (YC W26) – An open-source IDE for thoughtful software design
Show HN: Whiteboard (YC W26) — open-source desktop IDE where humans and agents co-architect software, not just co-write code. Early thread at 45 points.
AI Safety Is Mostly a Sex Cult in Berkeley
A critique of the Berkeley AI safety scene (19 points on HN).
Sam Altman’s remarks at the United Nations Security Council
OpenAI's take on global security risks and governance.
Gemini CLI 0.61.0
Gemini CLI 0.61.0 hardens sandbox filesystem boundaries and blocks indirect prompt injection via build file modifications and untrusted flags — update before running it on repos you don't fully trust.
Factory 0.226.1
Factory 0.226.1 — Droid now finishes replies while background tasks keep running instead of waiting, and Droid Shield can fix flagged files and rerun blocked commands when you ask.
Amp Shared Runners
Amp now lets you share a runner with your workspace via `--share` — point the team at a GPU box, Mac builder, or network-specific dev box and spawn agents on it from ampcode.com.
CryptTrace OSINT Toolkit
Crypttrace is an OSINT toolkit that traces crypto fund flows across Ethereum, Bitcoin, Tron, and Solana, labeling known entities and scoring risk. It is notable for its practical multi-chain forensics workflow and local interactive fund-flow graphs.
MetaMask Desktop
MetaMask Desktop is a cross-platform desktop app for managing a cryptocurrency wallet and interacting with Web3 apps. It is notable as the official desktop alternative to MetaMask's browser extension, offering improved stability and system integration.
Laya Decision Engine
A non-autoregressive System 1 decision engine for typed choice, score, and yes/no decisions over any text in 100+ languages in a single forward pass. It uses RLCD training and a router to select the right checkpoint per request.
FluxDown Download Manager
FluxDown is a Rust-based multi-protocol download manager supporting HTTP, FTP, BitTorrent, and HLS/DASH streaming with browser integration. It is notable for its IDM-style dynamic segmentation and up to 10x faster downloads, free and open-source.
Kev Decision Models
A family of small Jev-like decision models built on Qwen3.5/3.8 that users can train and run themselves. Its API matches TypeSafe's System One, enabling yes/no, multiple-choice, and rating questions in a single request.
XERJ AI Search
XERJ is a local AI search tool that autoindexes code, docs, logs, and PDFs for agents and RAG. It is notable for using 40x fewer tokens than grep and being Elasticsearch-compatible, reducing coding agent output tokens by 2.7x in a controlled study.
Ultrafast Web Agent
jev-ultrafast is a browser agent that uses a dynamic, indexed action space to execute web tasks with minimal text generation. It is notable for being positioned as the fastest and cheapest web agent, demonstrated by completing a flight search in 7.1 seconds.
Evidence Based Life Guide
An evidence-based life guide in Chinese sorted by cost-effectiveness, covering longevity, first aid, finance, legal issues, employment, insurance, relationships, and entrepreneurship. It includes 612 recommendations with cost, benefit, evidence level, and sources from journals and official documents.
SemIf OpenJev
SemIf-OpenJev provides semantic decision-making from open models that runs on a 3090 GPU at home. It is notable for being an independent, open implementation with a browser demo and no waitlist, replicating Jev-like functionality.
Nulltrace Privacy Toolkit
Nulltrace is a local-first privacy toolkit that maps personal data against 400+ brokers, queues CCPA/GDPR deletion requests, and checks for password breaches. It is notable for keeping all data on the user's machine without requiring an account.
Compartment Agentic Memory
Compartment is an encrypted, fully offline memory vault for AI agents that works with tools like Claude Code and Cursor. It is notable for providing persistent, cross-session agent memory with no network or API requirements.
Echo Dot Linux
techo5-dot replaces the Amazon Echo Dot's Fire OS with Alpine Linux, turning it into a local Home Assistant voice satellite. It is notable for eliminating Amazon's cloud and Alexa while preserving all seven microphones.
Cinekive Cinematic Archive
Cinekive is a local desktop app for filmmakers that uses semantic search to find cinematic frames by mood, technique, or director. It is notable for combining computer vision with a privacy-focused, local-only architecture.