Update documentation to enhance clarity on native language and skill generation
ci / server (push) Failing after 3m37s
ci / client (push) Successful in 14s

- Revised `AGENTS.md` to specify that the same seed, map, name set, and native language must produce the same roster in tests.
- Updated `protocol.md` to clarify that everyone in the school is generated with the selected native language, and related languages may appear at a low skill.
- Enhanced `ai.md` to explain skill growth mechanics based on state and the introduction of minimum ranges for skills not yet acquired.
- Improved `people.md` to detail the generation of skills and traits, emphasizing that not every person receives every skill and the implications of body attributes on skill acquisition.
- Adjusted `projects.md` to reflect the inclusion of native language in roster generation, ensuring comprehensive documentation of project components.
This commit is contained in:
Leonid Pershin
2026-08-19 22:27:20 +03:00
parent 5a00ad7746
commit a441ed9763
5 changed files with 51 additions and 23 deletions
+2 -1
View File
@@ -186,7 +186,8 @@
Осознанно мало:
- **Навык растёт на уроке** — от предмета, с поправкой на черты и на то, в каком человек
состоянии. Голодный учится хуже.
состоянии. Голодный учится хуже. Если ключа ещё нет (химия в восьмом), рост начинается с
`Range.Min`, а не с нуля «как будто навык был всегда».
- **Опоздание и прогул видны.** В кабинете 5Б на математике не двадцать три человека, а
двадцать один, и карточка каждого говорит, где он.