StudyBuddy Learning Companion

The Problem

Many learning tools turn teaching into forms, progress bars, and fixed courseware. When a user asks to “understand X,” “explain why X keeps confusing me,” or uploads a slide screenshot, those tools may either require onboarding first or dump a long standard answer. StudyBuddy targets a narrower situation: the learner already has a concept, a teaching material snippet, or a path-planning question, and wants an explanation that is understandable, checkable, and reusable.

How It Works

• Explicit triggering: it engages for learning intents such as “teach me,” “explain X,” or “what is X for,” for uploads like PDF, PPT, Word, or screenshots, and for flashcard or learning-path requests; casual chat, pure debugging, or non-learning discussion stays outside scope.
• Automatic mode selection: single-concept questions enter a quick mode that explains one point and produces one card; systematic learning or uploaded material enter full mode, covering initialization, diagnostics, cognitive adaptation, teaching, card drafting, and archiving.
• Stable explanation structure: content follows problem, conclusion, analogy, principle, application, and misconception, while the output remains continuous prose rather than a labeled checklist; terminology appears in the principle stage instead of opening with definitions.
• Dynamic style adaptation: it adjusts brevity, tone, humor, analogy density, and information density from user feedback, and can choose among tech-learning, study-notes, and minimal-business flashcard styles.
• Confirmable knowledge output: after teaching, it first generates structured flashcard drafts. Users confirm or revise for up to two rounds before content is locked and the host image-generation tool is called; results can be archived as local .doc documents or Feishu docs.

Boundaries

StudyBuddy fits concept explanation, material-guided reading, path planning, and review-card generation. It is not intended for pure debugging, system configuration, or non-learning discussion. Feishu output depends on the host having the relevant MCP connection, and image quality still depends on the host image generator; the skill checks text rendering, content consistency, composition, style, and detail quality before download.