News
AI, cybersecurity, machine learning, scraping & programming news.
RCTF – a platform for hosting cybersecurity CTF competitions
A Platform Built for CTF Organizers rCTF is a platform designed for hosting cybersecurity capture-the-flag competitions. The project addresses a common pain point for CTF organizers: deploying and maintaining a competition infrast...
"Torturing" LLMs in a Robot Prison Has Triggered the Dumbest Debate in AI Yet
GitHub project puts locally hosted LLMs through Saw-style experiments A GitHub project has sparked heated discussion on X after its creator began running torture and pain experiments on a series of large language models run locall...
Show HN: Solvi 1.0 – an LLM proposes, checks decide, every decision replays
A runtime where code decides and models only propose solvi is a decision runtime built on a simple split: plain Python functions and checks describe what to compute, while a model — whether an LLM, a custom classifier, or a small...
New Archestra's OpenAPPA Saturates Two Major Security Benchmarks with a 0% Attack Success Rate
Archestra has released OpenAPPA, an open-source security engine designed to prevent data exfiltration arising from prompt injection or model hallucination. The engine operates outside the agent's prompt and execution loop, providi...
Show HN: AgentiLoop Agent Mac GUI Agent Loop for macOS 14.6 or Later
Archestra has released OpenAPPA, an open-source security engine designed to prevent data exfiltration arising from prompt injection or model hallucination. The engine operates outside the agent's prompt and execution loop, providi...
Show HN: Cliproxy-rs, a Rust port of the CLIProxyAPI LLM subscription proxy
cliproxy-rs is a small server you run on your own computer that lets coding tools such as Claude Code, Codex CLI, or Cursor use the AI subscriptions and API keys you already have, like a Claude Max plan or a ChatGPT plan, through...
Merging With An LLM: Can an LLM reconstruct my values from small data?
LLM capabilities are increasing. They create risks for financial and logistical infrastructure, and make engineered pandemics more likely. Work on model safety and societal mitigation efforts follows. Perhaps even a pause enters t...
OpenAI safety leader quits, warning AI company's culture is 'broken'
A safety leader at OpenAI has quit the company, warning that its culture was broken and that AI firms were not "being nearly careful enough" about developing the technology. David Robinson, who led the writing of safety reports th...
Manage end-user OAuth consent for AI agents with Amazon Bedrock AgentCore
The user wants me to rewrite an article from AWS about Amazon Bedrock AgentCore's Consent portal into a substantial original article for their news section. Let me understand the requirements: 1. Structure: ~700-1
The Case for Bringing Your LLMs Home
For years the default answer to "which model should we use" has been a cloud API. Call an endpoint, pay per token, let someone else worry about GPUs. That default still works for plenty of workloads. But as enterprise AI usage has...