Research Planner

Round 2 — Goals & Outputs

• What would a successful outcome look like? What should you be able to say or do after this research?
• Who is the intended audience, and what decisions might they make based on your findings?

Round 3 — Scope & Constraints

• Are there specific papers, authors, methods, or time periods you know you want to focus on?
• Are there topics, methods, or papers that are explicitly out of scope?

Round 4 — Specific Hypotheses & Questions (optional, based on earlier answers)

• Do you have specific hypotheses or relationships you want to test or verify?
• Are there methodological debates or open questions in this domain you want to resolve?

Step 3 — Synthesize & Confirm

Summarize back to the user what you've understood:

• Their primary research goal in one sentence
• 3–5 candidate primary questions you plan to write
• Key thematic areas you plan to list under Areas of Interest
• Any scope boundaries to add

Ask: "Does this capture what you're looking for, or should I adjust anything?"

Step 4 — Write research-questions.md

Once confirmed, overwrite research-questions.md using the template below. Be specific and actionable — primary questions should be answerable by reading the documents in docs/.

# Research Objectives

<1–2 sentence summary of the research goal and context>

## Primary Questions

1. <Specific, answerable question drawn directly from the user's goals>
2. <...>
3. <...>
(aim for 4–8 primary questions)

## Areas of Interest

- **<Theme 1>**: <brief description>
- **<Theme 2>**: <brief description>
(aim for 4–6 thematic areas)

## Scope Boundaries

- **In scope**: All Markdown documents in `docs/`
- **In scope**: <any specific papers, methods, or concepts the user named>
- **Out of scope**: <any explicit exclusions the user named>

Step 5 — Suggest Next Steps

After writing the file, tell the user:

• What the Ralph Orchestrator will use these questions for
• Suggest running the orchestrator: "You can now run the Ralph Orchestrator to start building notes from these questions."

Quality Criteria

• Primary questions are specific and answerable from the docs, not vague ("What is DiD?" not "What are econometrics?")
• At least 4 primary questions written
• Areas of Interest map onto actual themes in docs/
• Scope Boundaries are concrete, not just the default placeholder
• The written file is valid Markdown with no broken formatting