https://www.microsoft.com/en-us/security/blog/2026/05/28/typosquatted-npm-packages-used-steal-cloud-ci-cd-secrets/
| website | https://xeraa.net |
| https://twitter.com/xeraa | |
| https://www.linkedin.com/in/philippkrenn/ | |
| github | https://github.com/xeraa |
| website | https://xeraa.net |
| https://twitter.com/xeraa | |
| https://www.linkedin.com/in/philippkrenn/ | |
| github | https://github.com/xeraa |
more cursor skills in the wild — here and everywhere else. spin up an elasticsearch cluster, manage kibana, configure security, set up OTel,...
and almost hidden under it is also the brand new MCP endpoint for the elastic docs
BM25: "why won't you die?!"
> the lexical retriever BM25 with appropriate setup outperforms neural rankers in most cases; notably, gpt-oss-20b with BM25 on the passage corpus achieves the highest answer accuracy across all retrieval settings in our study; BERT-based learned sparse and multi-vector dense retrievers generalise better than LLM-based single-vector dense retrievers; and re-ranking remains highly effective
BM25 just keeps hanging on. more in this new paper: https://arxiv.org/pdf/2602.21456
is it an API key for a cloud account or the cluster?
way too common of a confusion. so why not both?! released today for serverless projects and your elastic cloud account
PS: please be extra careful in your day to day use with these. they have an extra wide blast radius 💥
obligatory GTC keynote tweet when you make the top slides: building HNSW graphs with up to 12x the throughput and 7x faster merges on elasticsearch and NVIDIA cuVS (a GPU-accelerated library for vector search)
powered by CAGRA (graph-based ANN algorithm built to run natively on GPUs) that still works with CPUs for search
full post: https://www.elastic.co/blog/elastic-nvidia-cuvs-integration
and find us at GTC — we have a booth and I'll be around today and tomorrow
PS: yeah, this is a lot of acronyms
nvidia GTC where every second word (spoken and on booths) is AI, agent, or token 😬
PS: who else is around?