AI reading guides

Use this collection to plan a source-first summarization workflow. Start by deciding whether the result is a reading preview, a project note, or a comparison. Confirm the input before asking for output, and keep a stable label for each source. The summary should make reading easier without hiding where its claims came from.

Review the important claims against actual passages, including qualifications and meaningful omissions. Keep source fidelity separate from tool safety: an accurate summary does not prove that a service needs access to your private library or permission to take actions. The guides below cover task design, careful verification, and the boundary between untrusted webpage content and your own instructions.