Study and review with AI
Keep a learning route, teach-back practice and later reviews in one project
Study and review with AI start with your goal, existing knowledge and available time. The study-coach skill proposes stages, identifies prerequisites and asks you to explain a concept through its definition, rationale, application, example and boundary or counterexample.
Feedback targets logic gaps, ambiguity and contradictions. Local project records support resuming and reviewing, but a planned topic is not automatically a learned one. Internal ratings and review intervals are operational heuristics, not formal grades, mastery certificates or proof of an optimal learning strategy.
What it does
- Identify prerequisites before tackling material
- Practise explaining concepts in your own words
- Check due reviews when resuming
- Record progress and the next action locally
- Keep several learning projects separate
- Revisit explanations that remain unclear
Install and use
- Prepare the full bundle in a file- and command-capable agent with local Python, NumPy, sentence-transformers and a writable project location. Initial setup downloads the local embedding model from Hugging Face and requires internet and storage.
- For example: “Plan introductory linear algebra for three half-hour sessions weekly, checking prerequisites first and keeping a separate project.” State your goal and starting point.
- Explain concepts using the requested five parts, then revisit the flagged portion. Ask the agent to report the stage, open questions and whether records were saved successfully.
- Select the correct project when continuing and check due reviews. Treat incomplete storage or retrieval as unresolved; file existence does not establish learning.
FAQ
Source
Made by MingMaple. Original skill: study-coach, published on GitHub under the MIT licence.
View on GitHub