Editorial Policy
This site generates a large number of pages from a keyword dataset. The editorial rules below are what keep that scalable without turning into low-quality filler.
Keyword selection and filtering
We start from a search-demand export, remove junk rows and duplicate slugs, then apply an entity-level safety filter. Slurs, explicit sexual content, religious and alcohol/drug vocabulary, self-harm and abusive language, and ambiguous borderline terms are removed. Every removal is written to a CSV audit with a reason code, so the boundary is inspectable rather than arbitrary.
No invented claims
We do not publish fabricated statistics, fake user profiles, invented leaderboards, unverifiable endorsements or claims of formal certification. Practice stats shown on the site are real numbers stored locally in the visitor's own browser.
How a page is written
Each page is composed from a hand-written entity record (handshape, movement, palm orientation, memory cue, common mistake, usage note) and hand-written prose keyed to the query's intent. Related questions are answered from that record. Pages are not created by swapping a keyword into a single template sentence.
Clusters and internal links
Keywords are grouped into semantic clusters that appear as categories on the homepage. Every page links to its cluster hub, to sibling signs in the same cluster, and to the previous and next page in its cluster, so no page is orphaned.
Corrections
Correction reports are checked against standard ASL references before the entity record is updated and the affected pages are regenerated. We do not silently change a description without updating the source record.