Review a fiction draft with AI
Read a chapter range from several simulated perspectives. Each observation points back to the paragraph that prompted it, so you can assess it in context.
Review a fiction draft with AI when knowing the ending makes it hard to judge what the opening actually gives away. The reader-panel skill assigns different questions to separate simulated readings. A first-page reader considers whether to turn the page. A continuity reader checks what a character could know at that point.
Pick a range that ends before the reveal. The resulting notes quote enough of a paragraph to locate the problem and stay attached to that version of the manuscript. When you revise the chapter, you can check whether the original concern still applies.
What it does
- Locate the paragraph where an opening stops raising an interesting question.
- Check whether the selected chapters deliver the experience the genre promises.
- Track a detail that changes before the story explains why.
- Flag passages where an informed human reader could offer a perspective the simulation lacks.
- Compare notes from a second pass with the version that produced the first set.
Install and use
- Install the skill in an existing Story Skills project with drafted chapters. The full workflow uses the official Story CLI and the maker’s feedback-triage companion.
- Ask for a first-page and continuity reading of chapters one to three. Keep later chapters out of each reader’s context.
- Open the passage behind each note before selecting revisions. Keep simulated and human feedback in their respective rounds.
FAQ
Source
Made by Daniel Dewhurst. Original skill: reader-panel, published on GitHub under the MIT licence.
View on GitHub