All products

Build a learning path with AI

Choose verified study material, learn a section yourself, then teach it back. Turn that discussion into concise teaching text while keeping evidence of learning separate.

Build a learning path with AI around material you will actually study, rather than a complete course generated in advance. With systematic-learning, the agent researches the route and sources, you study a module, and you explain it back. Misunderstandings are checked before your explanation is reorganised into concise teaching text.

The initial folder contains a preface, a learning record and empty chapter directories. Module text appears progressively after real teach-back. Reading, explanation, verified understanding, factual checking and document completion remain separate states. That distinction makes it possible to resume a course without treating a polished file as proof that you learnt its contents.

What it does

  • Compare sources against the agreed learning scope.
  • Record prerequisites and exact study locations.
  • Use teach-back to reveal missing reasoning.
  • Create concise module text from actual discussion.
  • Resume from a recorded stopping point.
  • Revisit previous mistakes with new answers.

Install and use

  1. Install the full package in a web- and file-capable agent. Specify authorised sources, a writable course folder and your language; conversation otherwise defaults to Chinese. Local helpers use Python, with an equivalent file-tool fallback documented.
  2. Ask, for example, “Plan an introductory probability course using accessible free sources. I will study each module and explain it back before you write it up.”
  3. Study the indicated section and teach it back. Explain an important correction again rather than merely accepting revised prose.
  4. Review the module, learning record and preface links. Keep the course location for your next session.

FAQ

Source

Made by AliothChen555-star. Original skill: systematic-learning, published on GitHub under the MIT licence.

View on GitHub