We are working on making room for lexicons as a resource type beside corpora. The workflow will be very similar:

1. Create a lexicon resource
2. Create a config (also YAML, but different schema)
3. Upload sources (JSONL https://jsonlines.org/ as first step)
4. Run pipeline (will not do something like annotation, but will build files for Karplabb)
5. Install to Karplabb https://spraakbanken.gu.se/karplabb/

Later features include allowing more source formats and exporting for editing in Karp https://spraakbanken.gu.se/karp/