mail: send via Brevo's HTTPS API, since HF Spaces blocks SMTP ports

co-writer: enforce the word limit and flag directives typed into the draft

co-writer: strip JSON punctuation from truncated suggestion output

co-writer: reject suggestions citing figures not in the draft

grammar: name the explanation language explicitly

grammar: supplement LanguageTool with an LLM pass

Refuse non-prose input in verdict and generation tools

fix: give language detection enough tokens to answer

fix: translator auto-detect returned an empty translation

deploy: QA bug-fix batch (BUG-001..047)

Improve GPT-OSS writing speed and reliability

Fix CoWriter lint formatting

Deploy document workspace and CoWriter API

Remove deployment staging files

Add model choices and fix tone analysis

Refresh Anovo model routing to GPT-OSS

Deploy account deletion endpoint

Allow concise boilerplate cleanup

Improve production humanizer quality

Deploy contextual paraphrase refinements

feat: remove history router to reduce storage

fix: lint cleanup (unused imports, line length)

feat: add more premium models (Llama, Phi-4, DeepSeek, Cohere)

fix: handle Neon PostgreSQL idle connection drops

fix: update model list to verified working models

fix: PostgreSQL support, remove ephemeral SQLite /data path

clean: remove pycache files, add gitignore

force rebuild: Dockerfile v3, .dockerignore, updated README

trigger rebuild

fix: persistent DB at /data, health diagnostics, latest code

feat: model selector, admin portal, auto-migration, improved humanizer

feat: premium tier with GitHub Models 405B + prompt tuning

feat: overhauled humanizer + doc upload endpoint