some automation for claude code for managin a skill folder that I can select from

nvidia accelerated vllm with jina v5

dockerization of nova act web agent

using rank_eval and deepeval to improve search and rag example

script to poll tempest weather station

Some examples combining streamlit, es, and langchain

web proxy for hugging face inference endpoint

Messing with langchain, huggingface transformers, and Elasticsearch

encapsulate llm model in a REST API in a docker container

Messing around with creative commons open5e json files