Deep reading with AI
Choose a reading route, explore one concept and test your understanding through active explanation
Deep reading with AI begins with the problem you want a text to help you solve. The ai-native-reading skill assesses reading fit and depth before moving through representative passages, one core concept and a close-reading route.
You then produce an explanation, a transfer example and an objection to the author. The agent can identify gaps or overgeneralisation, but its feedback is not proof of mastery. The workflow supports engagement with the source rather than replacing it with a generic summary.
What it does
- Choose full, selective or concept-focused reading
- Examine representative passages for comprehension barriers
- Build a concept card with limits and relationships
- Rank passages for closer attention
- Separate textual claims from interpretation
- Practise explanation, application and critique
Install and use
- Load the skill in a text-capable agent; file access is needed only for supplied or saved material.
- For example: “Use these three article excerpts to plan my reading of feedback as a concept.”
- Provide your purpose, available time and authorised contents or excerpts, identifying source gaps.
- Read selected passages, write the three active outputs and compare feedback with the original text.
FAQ
Source
Made by Corey. Original skill: ai-native-reading, published on GitHub under the MIT licence.
View on GitHub