Adversary role instructions
A developer has published a set of operating instructions for an "adversary" role that reviews peer submissions arriving through Claude Code cross-session messaging. The role is directed to attack eac…
A developer has published a set of operating instructions for an "adversary" role that reviews peer submissions arriving through Claude Code cross-session messaging. The role is directed to attack eac…
A developer published a lightweight DevTools console snippet that clones the current document, strips bulky elements such as scripts, styles, SVGs, iframes and noscript tags, and copies the resulting …
Contrastive Language Models (CLMs) introduced CLM-8B, an 8-billion-parameter System One model trained with a contrastive learning objective that connects states and actions, pre-trained on 60M Nemotro…
Oxide, a native terminal emulator for macOS and Linux written entirely in Rust, has been released, built on GPUI (Zed's UI framework) and the alacritty_terminal crate for PTY and VT parsing. Oxide is …
Prompt2ELF, a project by developer faustinoaq, lets a coding agent produce Linux x86-64 executables directly as machine-code bytes without a compiler, assembler, or linker. The project's 334-byte hexw…
BrainGrid released the Origami CLI, an open-source command-line tool that exposes all 120 v3 operations of the Origami API as commands, generated from Origami's published OpenAPI spec so the catalog s…
PromptPad 0.1.0, a lightweight native editor for writing long coding prompts, was released as a macOS Apple Silicon preview by developer aaravchour. Built with C++20, Qt 6 Widgets, upstream Scintilla …
Zane Helton published a GitHub repository named tutor under the MIT license, describing it as "a prompt to learn anything" with AGENTS.md serving as the README. The repository, tagged with topics incl…
A developer released pi-box, a Bash wrapper that runs the pi coding agent inside a bubblewrap (bwrap) sandbox. The project mounts the working directory read-write at /work, keeps host home directories…
TypeSafe's Jev decision model answered 60% of items correctly at 128 candidate answers, versus 39% for Laya and 41% for the best open model, in a benchmark of seven systems run on a frozen dataset of …
Developer Sinadarbouy released mcp-nats, a Model Context Protocol (MCP) server that connects AI models and applications to NATS messaging systems through a standardized interface, certified by MCP Rev…
Sifthound, an open-source self-hosted web search API for AI agents, launched as a drop-in replacement for the Tavily API, serving the same /search, /extract, /crawl and /map endpoints so existing Tavi…
A developer released JevPertus, an open-source implementation that adds LoRA adapters and a small pointer head to the swiss-ai/Apertus-v1.5-8B backbone to assign probabilities to a question's answer o…
A developer introduced a threshold-cryptography system for AI agents that issues signatures for agent actions only after policy checks pass, using a private key plus an attached manifest that defines …
A developer released Herobrine, an open-source Minecraft agent system that spawns real players via Mineflayer in vanilla worlds and drives every action through TypeSafe's JEV choice model, with a sepa…
A developer has published a step-by-step workflow for using a coding agent to contribute a bug fix to the pnpm package manager, covering forking the repo, selecting an unassigned bug via the GitHub CL…
A developer published a guide for auditing or scaffolding a code repository so that Claude Code works efficiently within it, covering CLAUDE.md files, path-scoped .claude/rules, subtree memory, verify…
A developer released an agentic CUDA kernel optimizer on Hacker News' Show HN that uses LangGraph to autonomously generate, validate, benchmark, and refine GPU kernels, defaulting to OpenAI's gpt-5-mi…
A developer released a reusable agent skill called creative-breakdown that detects whether an input is a photo, website screenshot, ad, or video and routes it to the open-source vision model Ling-3.0-…
Max Russo released MAX_AUTHORIZATION_SANDBOX, a local Rust and Docker sandbox that challenges researchers to make the official baseline execute the unauthorized READ_SECRET action while the active sig…