License Compliance Notes

This repository includes AGPL-licensed upstream material and is licensed as AGPL-3.0-only.

What is included in-repo

assets/dictionaries/words-*.json are imported from keybr.com and tracked in THIRD_PARTY_NOTICES.md.
assets/dictionaries/words-<lang>.json.license records source and license for each imported dictionary file.
assets/dictionaries/manifest.tsv maps language keys to imported files/sources.
assets/dictionaries/SHA256SUMS stores dictionary checksums for integrity verification.
scripts/validate_dictionary_manifest.sh validates manifest entries, sidecars, and checksums.
scripts/derive_primary_letter_sequences.py derives per-language primary-letter sequence seed data from dictionary frequency.
assets/dictionaries/primary-letter-sequences.tsv stores the current derived output.
docs/unicode-normalization-policy.md documents NFC normalization policy and equivalence expectations.

The clones/ directory is gitignored and used for local analysis only.
References in docs/plans/ to third-party projects are primarily idea-level research unless explicitly documented as imported content.

Code drills can download source files from upstream GitHub repositories.
Passage drills can download texts from Project Gutenberg.
Downloaded content is cached in user data directories by default, not in this repo.
Downloaded code snippet caches include a *.sources.txt sidecar with source URLs.

If you import any third-party file into the repository, add it to THIRD_PARTY_NOTICES.md.
Add a sidecar filename.license (or equivalent) with source and license.
Keep the project license compatible with imported copyleft obligations.
If you later commit downloaded snippet caches or passage corpora, include attribution and the relevant upstream license terms for those files.