d8c9fa89959bfe6b3f68425f09e1ca8464154931
The comment instruction told the model to sweep up leftovers: "clinical notes, in the language spoken. Omit the parts already captured as treatment type, teeth, prosthesis work or deadline." So filler, small talk and an unrequested diagnosis all became a persisted clinical note — and comment is the one durable trace a recording leaves (§10). The model now reports the exact words that asked for the note in a new commentTrigger field, and comment holds only what was dictated after them. resolveVoiceIntent keeps comment only when a trigger is present, so the rule is enforced by code instead of trusted to the prompt: a model that decides leftover speech was a note loses that note. Per-locale trigger vocabulary goes in the prompt's locale notes, beside the tooth vocabulary — «بنویس که», "write this in the notes", "noteer". The resolver only checks that a trigger was reported, so it needs no per-locale knowledge and stays locale-neutral, as §6 requires. Also adds HARD RULE 7: never decide something is a note; speech that fits no field is simply not reported. Four resolver tests. The two that drop an unasked note fail without the guard; the trigger-only and trigger-with-note cases pass either way and exist to prove the guard does not swallow a real note. Frontend untouched — it only ever saw the resolved comment, never the trigger. Spec: §5 gains the rule and the field, decision 54 added. Gates: backend 220 tests, nest build, ESLint clean on the voice module; frontend tsc --noEmit clean, 52 Vitest tests. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Dyolink
Monorepo: NestJS backend (backend/), Next.js frontend (frontend/), Docker stack under infrastructure/..
Local development: see backend/README.md and frontend/README.md.
Cursor AI: project conventions for agents are in AGENTS.md, .cursor/rules/, and .cursor/skills/.
Production deploy (https://nudentic.ir)
Full guide: infrastructure/DEPLOY.md
Git tag v1.0.1 → Gitea workflow .gitea/workflows/prod-tag-deploy.yml builds images (URLs baked for nudentic.ir), pushes to Gitea registry, SSH-deploys the Linux stack.
First-time SSL / Hub fallback: same DEPLOY.md.
Staging deploy (https://wixur.ir)
Full guide: infrastructure/STAGING-DEPLOY.md
Merge or push to master → Gitea Actions builds images → deploys to https://wixur.ir (Gitea: http://wixur.ir:3000).
Workflow: .gitea/workflows/registry-build-deploy.yml
Related paths
| Path | Role |
|---|---|
backend/Dockerfile |
API image |
frontend/Dockerfile |
Web image |
infrastructure/STAGING-DEPLOY.md |
Staging setup, CI variables, testing |
infrastructure/docker-compose.registry.yml |
Pull-only staging stack (registry images + nginx + postgres) |
infrastructure/deploy.registry.env.example |
Template for deploy.registry.env |
Description
Languages
TypeScript
97.4%
Shell
1.1%
CSS
0.8%
JavaScript
0.3%
Dockerfile
0.3%
Other
0.1%