01Log
Archive
The full log, newest first.
GCC Rust Frontend
gccrs is a GCC front-end for the Rust programming language, still in early development and not yet usable for real Rust programs. Its goal is to become a fully upstream GNU toolchain implementation, giving Rust an alternative to LLVM-based compilation.
Homomorphic Encryption Demos
HEIR is a development platform and compiler toolchain that converts existing models into fully homomorphic encryption-compatible versions. It supports multiple FHE schemes and backends, using MLIR to abstract encrypted workloads for privacy-preserving computation.
Selfie Systems Toolchain
Selfie is an educational system that includes a self-compiling C compiler, a self-executing RISC-V emulator, and a self-hosting RISC-V hypervisor. It is notable for teaching systems engineering through self-reference in a unified undergraduate and graduate project.
Grain Compiler Toolchain
Grain is a programming language that compiles to WebAssembly via Binaryen, providing a modern toolchain and CLI. It targets systems-relevant web applications, with an active development community and a guided learning path.
Coding Agents for Generalized Task and Motion Planning Problems
Asks how far a coding agent gets when asked to write a robot policy — tests LLM-generated code against generalized task and motion planning problems.
Show HN: AI·rete·RAG – a Rete rule engine decides, RAG explains why
Show HN: a Rete rule engine makes the decision, RAG explains why — aimed at teams that need auditable LLM decisions in lending, fraud, or clinical triage.
PUBG Ally: A Conversational Embodied Agent as an AI Teammate
A conversational embodied agent that plays PUBG as a human teammate — tests whether natural-language coordination beats scripted bots in a live game environment.
Agent-Editing World Model: Rethinking World Modeling for LLM Agents
AEWM rethinks world modeling for long-horizon LLM agents — the paper argues agents should edit a learned world model rather than re-derive state each step.
Cline: Cross-Origin WebSocket Hijacking in Cline Hub Dashboard (`/browser` endpoint)
CVE-2026-59723: cross-origin WebSocket hijacking in Cline Hub Dashboard's `/browser` endpoint — an attacker's page can take over the socket. Anyone exposing the dashboard beyond localhost should patch now.
Polymarket Trading Bot
A retired Python bot that market-made and arbitraged Polymarket sports markets by de-vigging sportsbook odds and hedging. Notable for its real on-chain trading history, finishing about $5k net on a public wallet.
Intraday Stock Scanner
A real-time intraday stock scanner for US equities that streams 1-minute bars and displays configurable alerts in a customizable browser dashboard. Notable for running locally with user-provided Alpaca or Schwab data, with no subscription or telemetry.
Options Gamma Dashboard
An open-source, self-hosted options analytics dashboard tracking dealer gamma exposure, max pain, open interest, and IV surfaces. Notable for storing historical snapshots, enabling analysis that live-chain-only tools cannot provide.
Chan Python Framework
An open-source Python framework implementing Chan theory for Chinese market technical analysis, supporting morphology and dynamics buy/sell point calculation across multiple timeframes. Notable for its practical scope and active development among Chinese traders.
A-Share Quant Framework
A collection of Chinese A-share quantitative tools including the free Ashare market data API, a free tushare token, and a MiniQMT plus XtQuant local trading framework. Notable for providing a low-cost, privacy-focused path to real-money automated trading for individual A-share investors.
AI Trading Assistant
A self-hosted AI market monitoring assistant for A-shares, HK, and US stocks that integrates the TradingAgents multi-agent framework for portfolio analysis and decision reports. Notable for combining real-time monitoring, holdings management, and multi-channel alerts with AI-driven investment reasoning.
AI Investment Studio
An open-source, AI-native trading analysis and automation system where users compose visual workflows for market signal analysis and portfolio actions. Notable for its workflow-based approach to building custom AI trading systems with approval and broker integration, though still early-stage.
Claude platform release notes — September 24, 2026
Claude API now bills pre-output refusals in bio/frontier_llm/reasoning_extraction categories — watch your costs if you hit these refusal types.
Codex 0.157.0
Codex 0.157.0 adds GPT-6 Sol and Luna (with Bedrock support) and auto-starts background servers for interactive sessions — new models to test, and less manual server setup.
Amp Less Noise
Amp collapses agent step-by-step logs by default — you see outcomes, not every intermediate step. Expand when you need to debug.
GoLive Skill CLI
Open-source Agent Skill and zero-dependency Node CLI for deploying agent-built products. Notable for detecting needs, planning changes, and applying them on your own accounts without telemetry or a backend.
Itsaplan Project Management
Self-hosted project management and issue tracking alternative to Linear and Plane. Notable for treating AI agents as teammates with full REST API, webhook, and MCP access.
SGLang Course
Official SGLang and Datawhale course on LLM inference, teaching concepts through building a mini-sglang and reading SGLang source. Notable for hands-on learning aimed at inference engineers, with English and Chinese versions.
AI Agent Memory
Git-native persistent memory for AI coding agents implementing Google OKF v0.2 in pure Go. Notable for sub-300µs BM25 search and 80% token reduction without external databases.
NeoHorse Self Improvement
TokenRhythm's open-weight model family for agent workflows, including 4B and 9B models for tool use and coding. Notable for exploring recursive self-improvement via agentic post-training with a routing harness.
Graftcode Development Layer
Integration-free development framework for calling methods across languages and runtimes as local dependencies. Notable for eliminating SDKs and client code, aiming to replace integration layers entirely.
Doop Design Canvas
Open-source multiplayer design canvas where humans and AI agents edit together live. Notable for streaming agent edits via MCP and rendering real HTML in sandboxed iframes.
Harkinian Pad
A native iOS/iPadOS port of Ocarina of Time via the Ship of Harkinian source port, using Metal rendering and touch controls. It is notable for bringing a full PC source port to Apple mobile platforms without requiring a jailbreak.
Mugshot Screenshot Testing
A screenshot testing library for Compose Multiplatform and Android previews that turns annotated @Preview functions into tests. It runs on the JVM without an emulator, reducing boilerplate and making visual regression checks part of unit testing.
DSH Mobile APK
An Android APK shell for DeepSeek Harness that combines a WebView UI with an embedded Termux runtime snapshot. It is notable for providing a one-install way to run a web agent with real bash execution on-device without requiring the Termux app.
Overlay Translator
An open-source Android app that translates on-screen text in games, visual novels, and manga by capturing the screen, running OCR, and overlaying translations. It is notable for working without root and supporting on-device OCR and offline LLMs.
DDC Decompiler
A Rust-based decompiler that converts Android DEX bytecode back into readable Java at scale. It is notable for its speed, processing large APKs in seconds, and for producing output that javac can verify.
Jev Chat Jarvis
A mobile dialogue copilot for Chinese chat apps like QQ, X, and Feishu that reads on-screen conversations, suggests replies, and fills the input box. It is notable for its non-invasive accessibility-service approach and early momentum among Chinese users.
Always Strong
A single-flash Magisk/KernelSU/APatch module that bundles TEESimulator-RS and PlayIntegrityFork to achieve Strong Play Integrity on rooted Android devices. It is notable for simplifying an otherwise manual, multi-module setup for device attestation bypass.
Agents API multi-agent orchestration
OpenAI's Agents API docs now cover multi-agent orchestration — patterns for coordinating multiple agents under one API, with Codex and workspace agents as building blocks.
AdaMem: Adaptive Memory Token Allocation for Soft Compression in Retrieval-Augmented Generation
Adaptive memory token allocation for soft compression in RAG — dynamically assigns memory tokens to compress retrieved context, directly relevant to long-context agent memory budgets.
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.