This version:
* Improves how the LLM enrichment report reads. The Markdown
structure the model returns is now rendered with color instead of
raw markers: the title and section headings (# and ##) and the
**bold** key terms and verdicts are highlighted, and the markdown
symbols themselves are removed, so the report reads as a formatted
assessment rather than plain markup. On a dark background the body
text is shown in a near-white tone for stronger contrast and
easier reading, while headings and labels keep their
blue/purple/cyan accents. The same rendering is applied
consistently across the CLI, the interactive REPL and the TUI, and
the enrichment prompts now ask the model to use that heading and
bold structure so the formatting stays consistent.
