Blocklists
All in src/enrichment/config.ts (~790 lines). Maintained as the result of 29 Autoresearch Loop rounds — every false positive observed in an experiment got an entry.
The lists
INVALID_NAME_STANDALONE_WORDS (~400 entries)
A Set<string> of lowercased single words that, if present in a candidate name, disqualify it. Categories:
- Navigation items:
kontakt,om,tjänster,produkter - UI phrases:
cookies,performance,analytics,cookie - Industry terms:
podcast,golf,event - Honorifics
- Common false positives observed in test runs
INVALID_SUFFIXES
Legal-entity suffixes that, if a name ends with them, mean it’s a company not a person:
AB, HB, EF, KB, LLC, Ltd, GmbH, förening, etc.
LOCATION_TERMINATING_WORDS
Compass directions and region labels: Syd, Nord, Öst, Väst, etc. Catches “Regionchef Syd” being mis-read as a person.
INVALID_DOMAINS (140+)
See Domain Blocklist.
Used by
- Name Validation — reads the first three lists
- Domain Discovery — reads
INVALID_DOMAINS - Crawlee Scraper — emails with generic local parts (
info,support,hej,kontakt) are filtered by a similar list
See also
Name Validation, Domain Blocklist, Autoresearch Loop.