AI Reader tools (2)

MCP isn't just read/write drafts — you can also trigger AI analysis from outside.

1 · list_personas

List available AI reader personas.

Input: (none)

Returns:

• Persona id / name / attributes (age / preference / culture / tolerance)

Same data as in-Slima persona selection.

2 · analyze_chapter

Run AI analysis on a chapter (not a full-book reading).

Input:

• book_token, chapter_path
• personas: array of persona ids (max N, per tier)
• analysis_type: enum (quick / standard / deep)

Returns:

• Score / observations / suggestions
• Same output as AI chapter analysis inside Slima

This isn't "run a full reading" — that's still only triggerable from Slima's UI (on the roadmap).

Example (Cursor)

You: use list_personas, what are my options?
Cursor: [lists]

You: use analyze_chapter on Chapter 3 with "fantasy fan 25" + "literary critic", quick mode
Cursor: [calls analyze_chapter] → report

Limits

• Daily quota tied to Slima subscription
• Chapters > 50,000 words exceed token cap
• Script Studio books: analyze_chapter doesn't support structured scenes (.scene)

See: Script Studio tool limits

Related

• File operation tools
• Slima AI Beta Readers
• AI chapter analysis