dsh-stddev

In model-inference and data-processing conversation flows, standard deviation is one of the most commonly needed measures of dispersion, yet you rarely want to pull in a full statistics library or fire off a network request just to compute a single number. dsh-stddev is a lightweight tool built for exactly that: it registers standard-deviation calculation as an in-session tool that the model can invoke on demand during reasoning, with zero extra configuration.

The entire plugin is implemented in pure Node.js and runs entirely locally with zero network dependencies. That means it stays stable in offline environments, intranet deployments, or latency-sensitive pipelines, and will never break because an external service is unreachable. Usage is equally straightforward: once installed, you simply call the tool the plugin registers within your conversation, with no API keys or additional dependencies required.

It is well suited for developers and teams who need instant statistical metrics inside a model-inference workflow and want to keep their dependency footprint minimal, avoid heavyweight scientific-computing libraries, or operate under network-isolation constraints. Released under the MIT license, it is free to use, fork, and extend for personal or commercial projects.