ドキュメント › AI Readers › Character analysis diagnostic

Character analysis diagnostic

最終更新 May 18, 2026 · 約2分で読めます

When readers say "I felt nothing reading this" it's usually a character issue, not a plot issue.

AI Beta Readers "Reader Feedback" tab: per-reader diagnostics — Opening Impact and Character Analysis — each list every reader (Alex, Olivia, Marcus) with their own 0-10 sub-score next to a short comment (illustrative content).

Sub-scores

The Character analysis diagnostic has four sub-scores:

Sub-score Examines
Protagonist Will readers spend a whole book with this person
Motivation Are the character's goals / inner needs clear and believable
Dialogue Is the dialogue natural, does it move the scene
Voice Do different characters speak / think distinctly

Each reader scores the character dimension (0-10) with a comment, and the consensus is often flagged (e.g. "Consensus 3/3: characters need improvement").

Common issues

  1. Protagonist too perfect → no entry point for readers
  2. Protagonist too miserable → readers withdraw
  3. Supporting characters are tools → no goals of their own
  4. All characters speak the same → dialogue is one author voice
  5. Placeholder-style names (e.g. J. Doe) → readers can't invest

Pairs with Editor detail

"Character Depth" is also one of the AI Editor's 5 dimensions — the Editor detail tab has more granular, "Mark as Resolved" suggestions.

Related

Slimaで試してみる

アプリを開いて、自分の作品で同じ手順をたどってみてください。無料で始められます。クレジットカードは不要です。

Slimaを開く
この記事は役に立ちましたか？