Version 0.1.4 release
Release 0.1.4 improves CLI output consistency and expands generation workflow
flexibility for local and OpenAI-compatible LLM providers.
Highlights
open-data-products explain <document> --jsonnow returns structured ODPS
JSON instead of embedding the line-oriented explanation as a JSON string.- ODPS explain JSON includes product metadata, component count, compliance
level, production-readiness status, and data access state. open-data-products summary <document>now prints a compact human-readable
metadata report by default.open-data-products summary <document> --jsonpreserves the previous
machine-readable summary payload.open-data-products --versionandopen-data-products -Vnow report the
installed SDK version.- Top-level
open-data-products --helpnow shows current generation examples. - Help examples now use project-owned config names,
generated/output folders,
provider/model overrides, and custom prompt folder usage. - Removed stale
generation.config.yamlexamples from the top-level help text. - Added
openai-chatprovider support for local OpenAI-compatible servers such
as LM Studio, vLLM, llama.cpp server, and LocalAI. - Local model names stay user-controlled strings; the SDK validates provider
shape and paths without enforcing a fixed model catalog. - Bundled generation config now includes editable local profiles for Ollama,
LM Studio, and vLLM.
Docs Updated
RELEASE_SUMMARY.md- CLI top-level help text
README.mddocs/generation.mdllms.txt