Developer Simon Willison has released version 0.1a9 of the datasette-llm plugin. The update introduces a user experience improvement by allowing the plugin to automatically pick up the default model used by LLM. Consequently, users can launch Datasette Agent using the command line tool without needing to manually configure a default model parameter.
Streamlined Command Line Workflow
Previously, running the tool required an explicit model specification flag alongside the plugin configuration. By inheriting the default model setting directly, the new release reduces friction for developers testing database interactions via the command line. This release serves as an incremental update ahead of larger planned upgrades for Datasette Agent, though specific details regarding those upcoming features were not provided in the announcement.
Unresolved Details and Future Updates
Aside from the core change regarding model inheritance, the announcement did not specify a full changelog or timeline for the subsequent larger upgrades to Datasette Agent. It remains unstated whether further configuration simplifications are planned for future pre-releases in the 0.1 alpha cycle. Users running the software should watch for future version releases to see how the broader pipeline upgrades unfold.
The release is available now for installation and testing through standard Python package workflows.
Lead image: Rafael Minguet Delgado / Pexels
