Seeking Feedback: WeKnora Knowledge to OpenMAIC Classroom Workflow

[peterchan2020]

Hi OpenMAIC community 👋

I’d love to share a new feature I’ve built for the OpenMAIC Classroom skill, and would greatly appreciate your feedback!

Key Updates:

1. Wiki Knowledge Graph-based Micro-Classroom Generation
   Added support for concept-level micro-lesson creation by traversing WeKnora Wiki knowledge graphs. It converts every concept page and its related entities into structured OpenMAIC requirements, enabling batch generation of targeted micro-classrooms.
2. Pure-text Concept-to-Requirement Converter
   Added a two-stage script (scripts/concept-to-requirement.py) that transforms concept graph materials into pedagogical design JSON → requirement strings. This is completely LLM/embedding-free, with clean entity sorting based on dependency order.
3. Updated Workflow & Templates
   • Refreshed SKILL.md with the full Phase 1.2 workflow for this new input source (Wiki read → script parse → sequential API call → manifest-based resume support).
   • Added Template 5 (Concept Graph) to references/requirement-builder.md, defining standardized input/output structure and entity classification rules.
4. Resumable Batch Processing
   Implemented checkpoint resume support to prevent large-scale classroom generation from failing unexpectedly.

PR Reference:

Tencent/WeKnora#1119

I’d especially value feedback on:

• The structure of the new requirement template for concept graph inputs
• Usability of the batch generation workflow
• Potential edge cases or improvements for the converter script
• Ideas to extend this to other knowledge base formats

Thank you all for building such an amazing ecosystem! Looking forward to your thoughts 🙏