01Log
Archive
The full log, newest first.
Aurelia Framework
Aurelia 2 is a standards-based front-end JavaScript framework for building browser, mobile, and desktop applications with minimal framework intrusion. It is actively maintained but has a smaller community compared to other frameworks.
React Font Awesome
Official React component that renders Font Awesome icons as SVG with JavaScript, providing an API for using Font Awesome in React apps. It is the essential, well-maintained wrapper for the icon library.
ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement
ModularRSI proposes modular recursive self-improvement for agent harnesses — generalizable across tasks, not task-specific.
Show HN: Pelican-bicycle alternatives (updated for 2026)
HN user tests 30 LLMs on SVG generation prompts inspired by Willison's pelican benchmark — updated for 2026 models.
Apple's Siri AI Can Be Swapped Out for Claude, ChatGPT, Code Shows
Code shows Apple's Siri AI can be swapped for Claude or ChatGPT — HN's 225-point thread debates Apple's closed ecosystem vs open model choice.
Benchmark Radar: A Living Database and Search Engine for AI Benchmarks and Evaluation
Benchmark Radar is a searchable living database of AI benchmarks — aggregates sources, score histories, and discovery for eval selection.
Mistral X Mozilla: Private, Multilingual AI Browsing
Mistral and Mozilla partner on private multilingual AI browsing — HN's 400-point thread centers on local vs cloud inference for privacy.
Anthropic SDK 1.6.0
SDK 1.6.0 exposes the compaction parameter and signed blocks (beta), auto-mode Managed Agents permissions, and async credential token providers.
Claude API: on-demand compaction beta + Managed Agents auto permissions
Messages API can compact a conversation on demand (beta `compact-2026-09-04`); Managed Agents permission policies add `auto` so the server evaluates each tool call.
Union Alpha
Multimodal model built for research, coding, and agentic workflows with frontier-level general performance.
Data Maskit
A local privacy gateway that masks sensitive data in requests to AI coding assistants and restores masked values in streaming responses. It supports tools like Cursor, Claude Code, Codex, Pi, and any configurable Base URL tool.
Compartment Deployment System
A self-hosted Kubernetes deployment system for running small software on infrastructure your team controls. It installs into an existing cluster or provisions a single-node Kubernetes host and provides runtime, URLs, access control, deployment history, and operations surface via repository-first workflow.
CI/CD Beginner Template
A production CI/CD pipeline template for Node.js projects, with a minimal Fastify service and quality gates, container security, SLSA provenance, cosign signing, and automated release management. It includes ESLint, Prettier, TypeScript checks, Vitest coverage, and Node 22/24 test matrix.
Meme Radar
A local, read-only multi-chain meme candidate scanner and manual review tool. It uses GMGN, GoPlus, and DexScreener to surface evidence on contract risk, liquidity, market cap, and official links across Solana, BNB, Base, Ethereum, Robinhood Chain, Arc, and Stable.
DeepSeek Harness Bridge
A plugin that adds multi-channel remote access and security controls to DeepSeek Harness, including LAN QR connections, Cloudflare or self-hosted tunnels, and WeChat/QQ/Feishu/Telegram bots. It provides protocol-wide authentication, anti-tamper protection, and disaster recovery.
Enterprise System Design
An enterprise distributed systems design course and technical reference for building reliable, secure, operable platforms on Azure, from first principles to production AI workloads. It spans enterprise, distributed, AI, cybersecurity, reliability, cloud, HPC, edge, and mission-critical infrastructure.
React Native Secure WebView
A Fabric-native React Native WebView for authentication, checkout, OAuth, and other controlled web flows. It denies by default, validates top-level navigation against an origin allowlist natively, intercepts custom scheme redirects, and exposes a single string-only web-to-native bridge.
LLM Token Optimizer
antiburn is a local desktop app that checks AI coding sessions for token burn causes like deep sessions, excessive model use, broken caching, and unused tools. It supports Claude Code, Codex, Cursor, Copilot, and others.
Markdown Workflow Runner
upmd converts Markdown code blocks into dependency-aware runnable tasks and workflows, executed in a real pseudo-terminal from TUI, CLI, or CI. Output stays beside source, and shell environment changes carry to later tasks.
Privacy Preflight Tool
ShareLint is a local, fail-closed privacy preflight that scans files, folders, and nested archives for secrets, PII, metadata, speaker notes, and hidden content before sharing, without uploading them. It is practical for preventing accidental data leaks.
Markdown Template Language
Knap is a simple template language that converts data into Markdown using variables, filters, and logic, and powers Obsidian Web Clipper and Importer. It executes templates without eval or arbitrary JavaScript.
TUI Connection Manager
ctty is a lightweight Go TUI connection manager handling SSH, serial, SFTP, Telnet, and FTP in one ~5MB binary. It offers GUI-like convenience without Electron overhead.
Binary Analysis Tool
knife is a Rust reverse-engineering toolkit that parses, triages, disassembles PE/ELF/Mach-O binaries and recovers functions/CFG, scans crypto constants and YARA. It merges five tools into one static binary and pinpoints dangerous call sites like memcpy length from subtraction.
TUI Music Player
Mousiki is a customizable terminal music player for local playback and online streaming, featuring spectrum visualizers, synced lyrics, and keyboard-driven controls. It is fast, lightweight, and Apache-2.0 licensed.
PTP Protocol Book
An open-source book on Precision Time Protocol (IEEE 1588/gPTP) covering protocol internals, LinuxPTP source code analysis, and a hands-on ~1000-line C implementation. It progresses from thought experiments to running code, aimed at readers without prior network protocol knowledge.
AI Embedded Toolset
Skill pack for AI coding assistants covering MCU firmware build, flashing, debugging, serial/protocol monitoring, and pipeline orchestration on Linux, macOS, and Windows; installation requires project analysis and user confirmation before copying skills.
ESP32 Pomodoro Timer
A Pomodoro timer for the Waveshare ESP32-S3-Touch-LCD-3.49 development board. It has a touchscreen interface, three customizable timer profiles for focus and breaks, and custom background support.
IC Design Documents
A Vietnamese-language document collection spanning basic electronics to IC design, focused on front-end RTL/design verification, with supplementary knowledge for other electronics roles. Authored by Phùng Quang Minh, it presents fundamentals in a friendly, systematic way without relying on shortcuts.
AI Embedded Skills
Open-source embedded development and debugging skills for Claude Code, Copilot, TRAE, and other Skill-protocol AI assistants, enabling direct operation of compilers, debuggers, and communication buses from code generation to hardware verification.
Connected Machine Runtime
An Apache-2.0 C runtime for constrained devices like ESP32-class hardware, providing on-device identity, attestation, and BLE-native stablecoin payment primitives. It keeps private keys on-chip, enabling devices to act as transaction parties rather than sensors represented by others.
ESP32 Hardware Sandbox
Sandboxed Lua runtime for ESP32 devices with network hot reload and hardware peripherals exposed through a driver interface, plus an optional managed websocket backend for remote app loading.
Gemini 3.8 Live and 3.8 Live Extended Thinking
HN thread on Gemini 3.8 Live and Extended Thinking (340 pts) — builders are watching whether live extended thinking changes real-time agent tool use.
When Agents Slow Down: Understanding LLM Agents' Test-Time Strategies via Elo-per-token Analysis
Elo-per-token analysis shows agents initially beat independent sampling but hit diminishing returns; parallel short trajectories may scale better.
Emergence World: Adversarial Stress-Testing of Long-Horizon Multi-Agent Systems
New benchmark adversarially stress-tests long-horizon multi-agent systems — a direct probe of where orchestration breaks under pressure.
ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents
ScienceBuddy proposes recursive self-improvement loops for interactive scientific agents — worth a look if you're building agents that refine their own tool use.
Harness or Model? Isolating the Harness Effect in Agentic Coding with a Contamination-Controlled Private Suite
Uses a contamination-controlled private suite to separate harness from model gains in agentic coding — tells harness builders what actually moves the needle.
Toward Self-Adaptive Physical AI: Can LLM Agents Manage Long-Horizon Physical Tasks?
Probes whether LLM agents can self-adapt across long-horizon physical tasks — relevant for embodied agent operators evaluating deployment readiness.
Antigravity — New Permissions System & Terminal/Git Version Control for Enterprise
Antigravity adds a new permissions system and terminal/Git version control for Enterprise — tighter access control and versioned terminal workflows for enterprise deployments.
Gemini CLI 0.60.0 — security hardening release
Gemini CLI 0.60.0 hardens MCP OAuth (RFC 9207 issuer checks), strips a hardcoded Google CrUX API key, and adds consent prompts for environment changes — security-focused release worth updating for shared environments.
WeVisDoc 4B
Apache-2.0 document vision model at 4B parameters, aimed at document understanding tasks.