v0.1.1

Changelog

• be52957 docs: show wenyan output token counts (Qwen ~42 vs OpenAI 67)
• 3cee50c refactor(wenyan): rename ultra-wenyan level to wenyan
• 6863a83 feat(wenyan): keep only ultra-wenyan; expand lexicon to ~380 words
• dda6adb docs: show all 3 wenyan modes on the PR-review example in the hero
• 0a48990 docs(wenyan): show the PR-review paragraph and the coverage limit
• 146ef3b feat(wenyan): join adjacent 文言 chars; add comparison table to README
• 473b72b fix(wenyan): yes -> 諾 (assent), no longer colliding with be -> 是
• 7201575 feat: wenyan modes — swap reduced words for Classical Chinese characters
• 31b99e3 fix: preserve URLs, paths, code, and versions verbatim through reduction