chinese-learning-orchestrator

Organize Mandarin pronunciation practice using an LLM-first pedagogy backed by a local xuezh engine, which runs mechanical transforms and produces audio artifacts for structured tone work and pronunciation feedback.

An LLM-first pedagogy for Mandarin instruction uses a local ZFC-style engine to enforce tool contracts and bounded context, ensuring deterministic, auditable interactions for review, graded input, and HSK audits.

Run an HSK-style audit by initiating a Mandarin lesson via the xuezh CLI, which generates bounded reports and logs structured interactions to review graded input and track learning progress.

Yes, the Azure pronunciation assessment pipeline integrates with the local xuezh engine workflow to generate audio artifacts and provide pronunciation feedback during structured Mandarin speaking practice.

Yes, you can use the local xuezh engine for repeatable Mandarin tone work because it enforces bounded context, clear tool contracts, and structured logging to ensure reliable, repeatable learning interactions.

The ZFC-style xuezh engine requires strict adherence to tool contracts and bounded context, meaning it handles mechanical transforms and deterministic interactions rather than open-ended conversational Mandarin practice.