The CLI brings two of Doxbrix's most useful capabilities to your terminal: full-text **search** across your content, and AI **content-quality** scoring. Both are...
The CLI brings two of Doxbrix's most useful capabilities to your terminal: full-text **search** across your content, and AI **content-quality** scoring. Both are...
The CLI brings two of Doxbrix's most useful capabilities to your terminal: full-text search across your content, and AI content-quality scoring. Both are scriptable — ideal for finding content and auditing quality in bulk.
Search content
dxb search runs the same full-text search your readers use, against your project's content:
dxb search "reset password" -p my-docs| Flag | Description |
|---|---|
| `--limit <n>` | Max hits (1–100, default 20). |
| `--locale <locale>` | Locale to search. |
| `--json` | Machine-readable output (global flag). |
# Find every page mentioning a deprecated term, as JSON
dxb search "legacy auth" --json | jq '.[].file'This is a fast way to find all references before a rename or deprecation.
Score content quality
dxb quality runs the same Scan with AI analysis from the terminal, scoring pages on readability, structure, SEO, accessibility, and completeness. It has the aliases dxb health and dxb score.
# Score every page (worst first)
dxb quality| Flag | Description |
|---|---|
| `--space <slug>` | Limit to one space. |
| `--sort <order>` | Page order: `worst` (default), `best`, or `title`. |
| `--min <0-100>` | Only show pages below this overall score. |
# Audit one space, worst pages first
dxb quality --space guides --sort worst
# Only show pages below 70
dxb quality --min 70The output mirrors the Docs Health board in the app — a prioritized list of what to improve.
Quality vs. lint
These check different things:
dxb lint— correctness: broken links, bad frontmatter, unknown blocks, orphans.dxb quality— how good the content is: readability, structure, SEO, accessibility, completeness.
Run both in CI — lint blocks broken docs, quality surfaces weak docs.