Crate blendtutor_core 

blendtutor-core

Domain logic and adapters for blendtutor: the lesson model, language runners (R and Python), grading, and LLM providers — built up slice by slice.

This is the reusable core a future Tauri GUI can call directly, with no process boundary. It deliberately does not parse command-line arguments or render terminal output; that is the responsibility of the blendtutor-cli crate, which depends on this one (the dependency only ever points cli → core).

Modules

course

Course discovery: a course directory’s manifest and the lessons it lists.

crypto

Password-based encryption for built sites: AES-256-GCM + PBKDF2.

eval

The eval-case model: a typed model and its parse boundary.

grade

The lesson↔runner grading join: pick a language’s runner and turn each lesson check code-string into a typed pass/fail outcome.

lesson

The lesson schema: a typed model and its validating parse boundary.

llm

The LLM feedback layer: a pure prompt builder and the rig-core provider client that turns a model’s structured tool call into a typed verdict.

quarto_export

Pure transforms: Lesson → Quarto .qmd fenced-div snippet or complete page, plus the static API key page (ADR-0019).

run

The student run loop: execute a submission, grade it, ask the LLM for a verdict, and bundle the result into a typed RunReport.

runner

The language-runner seam: a Runner trait and the normalized result of running learner code.

scaffold

Course scaffolding: the templates a new course starts from and the plan that writes them.

site

Static-site assembly: turn a course into a browser-deployable lesson site.

smevals_gen

smevals eval-dir generation: a pure translator from a lesson + eval suite to the file tree smevals run/build consume.

Structs

NotYetImplemented

Error for a command that is planned but not yet implemented in the current slice.