USB (Universal Systems Bridge) is an open-source tool for finding, installing, and moving AI skills and capabilities across different agent runtimes. It addresses two problems: skills and MCP servers are scattered across repositories and configurations, and users often have little way to check what they do before installing them. The project offers a catalog of 529 skills, a command-line interface, and an API, with support for 16 runtime targets. Its installer detects local runtimes and can create native files or portable bundles; users can search the catalog, install individual skills or presets, and verify downloads before running them.
A central feature is the Trust Layer, which adds checks beyond installation. The secscan command statically compares a skill’s stated purpose with capabilities indicated by its instructions and bundled scripts, such as accessing credentials, running processes, or deleting files. It reports evidence with file and line references, assigns verdicts and severity, and can block CI on new medium- or high-severity findings; it never executes the material being scanned. Baselines let teams distinguish new issues from known ones, while vague authority language is advisory rather than a standalone failure. For MCP servers, mcprobe observes startup behavior with operator consent but does not invoke tools. Together, the checks aim to expose mismatches between declared purpose and observable behavior, not to guarantee that a capability is safe.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Researchers tested 16 LLMs on all 55 TIS-100 puzzles, giving each one attempt and time to refine its code. The models found 133 new performance tradeoffs across 18 puzzles and set three outright cycle records, though human players improved on most solutions within days. (tis-100.wincon.dev)
Australia is considering a new criminal offence for AI companies that fail to take reasonable steps to prevent harm caused by their models. The government is also weighing an AI safety watchdog with power to impose significant fines, following a hack of a Medicare statistics portal involving OpenAI. (abc.net.au)
AI is increasingly becoming a gateway through which people find information. Audiences still resist having AI produce their news. (niemanlab.org)
Nvidia holds a central position in the AI industry, making disruption to the company potentially catastrophic for the wider ecosystem. The race for advanced AI is also intensifying economic and geopolitical conflict, while concentrated corporate power and infrastructure face growing political resistance. (france24.com)
LLM inference can require reading about 26 GB of model weights to generate each token from a 13-billion-parameter model. The explanation shows how model size drives memory bandwidth demands during text generation. (youtube.com)
Deletion Tuning matches full fine-tuning across three LLM post-training methods by zeroing fewer than 0.05% of model weights. Its compact masks also preserve prior-task accuracy better than full fine-tuning, LoRA, and random sparse tuning. (arxiv.org)
Today's best Hacker News stories, summarized and screenshotted, one email a day.