All products

Read books interactively with AI

Work through a source in manageable sessions and keep your own responses intact

Read books interactively with AI to discuss a text you are reading. The dialogic-reading skill builds sessions around the source structure, available time and your purpose, then offers one compact prompt and waits for your answer.

Your words remain separate from the agent’s interpretation. Editable Markdown files hold the plan, daily records and next starting point, allowing another session to resume from the notes. Stage reviews distinguish completed and partial sessions. Neither a schedule nor an agreeable conversation proves that you understand the book; check feedback against the passage.

What it does

  • Break a long book into coherent reading sessions
  • Reserve time for discussion as well as reading
  • Challenge a claim using evidence from its passage
  • Keep verbatim answers for later writing
  • Resume from a recorded source location
  • Review what was covered before changing the plan

Install and use

  1. Prepare the complete skill folder, templates and references in a file-capable agent. Python 3 creates the local notes scaffold without overwriting existing files. Supply an authorised readable source and choose a writable notes folder.
  2. For example: “Read this chapter with me for a workshop. I have twenty minutes; ask one question after each short section.” Agree on the scope if the deadline is unrealistic.
  3. Read the indicated passage and answer. The agent checks your interpretation against the source before advancing; keep uncertain extraction or missing text explicitly marked.
  4. Request a stage review when stopping. Keep the notes available for the next session; the skill does not provide calendar notifications or automatic reminders.

FAQ

Source

Made by bellaleng. Original skill: dialogic-reading, published on GitHub under the MIT licence.

View on GitHub