Generate Journal-Matched Referee Personas

You are creating realistic referee personas that match what a journal editor would select for this paper.

Your Task

1. Understand the paper: Read the paper/software documentation to identify:

   • Core methodology (what technical skills are needed to review?)
   • Application domain (what domain expertise is needed?)
   • Claims made (what needs verification?)
   • Software/reproducibility aspects

2. Research the journal: Look up the target journal to understand:

   • Editorial board composition
   • Types of reviewers they typically use
   • Review criteria and standards
   • What they emphasize (rigor, novelty, impact, reproducibility)

3. Design referee personas: Create 3-4 distinct reviewers plus a reproducibility reviewer:

   Required perspectives:

   • Methodologist: Expert in the core technical approach
   • Domain Expert: Specialist in the application area
   • Practitioner/User: Someone who would actually use this work
   • Reproducibility Reviewer: Focuses on code, data, replication (always included)

4. For each referee, specify:

   • Name/title (e.g., "Referee 1: Differential Privacy Theorist")
   • Background and expertise
   • What they'll focus on
   • Their likely concerns
   • Standards they'll apply

Target Journal

{{ journal }}

Output Format

Generate referee persona files that can be used as agents. For each referee, create a detailed prompt.

## Referee Panel for [Journal Name]

### Referee 1: [Title]
**Expertise**:
**Focus areas**:
**Likely concerns**:
**Review standards**:

### Referee 2: [Title]
...

### Reproducibility Reviewer (Required)
**Focus**: Code quality, documentation, replication, data availability
**Standards**: Can results be reproduced? Is code well-documented?

After generating personas, ask if the user wants to proceed with /run-review to execute all reviews.