Loading
Research Question [What are we trying to find out? Be specific — vague questions produce vague answers.]
Recommended by author
This prompt takes no variables — just pick a model and run.
# Research Project ## Research Question [What are we trying to find out? Be specific — vague questions produce vague answers.] ## Why This Matters [What decision or action does this research inform?] ## Scope & Constraints - Time horizon: [e.g. "focus on last 3 years"] - Geographic scope: [e.g. "US market only"] - Out of scope: [what we're explicitly NOT researching] ## File Structure ``` research/ ├── CLAUDE.md ← You are here ├── sources/ # Raw source material — one file per source │ └── TEMPLATE.md # Copy this for each new source ├── notes/ # Running notes, hypotheses, open questions │ ├── questions.md # Questions to answer (checked off as answered) │ └── findings.md # Key findings as they emerge ├── drafts/ # Work-in-progress documents └── output/ # Final deliverables (reports, memos, presentations) ``` ## Research Process 1. **Define** — clarify the question, agree on scope (this file) 2. **Gather** — collect sources into `sources/`, one file each with URL + summary 3. **Synthesize** — update `notes/findings.md` as patterns emerge 4. **Draft** — produce structured draft in `drafts/` 5. **Finalize** — clean up into `output/` deliverable ## Source Standards - Every claim needs a source — cite URL, date, and author where possible - Rate source reliability: Primary (official data) / Secondary (analysis) / Opinion - When sources conflict, flag it explicitly — don't pick a side silently - Prefer recent sources unless historical context is the point ## Output Format Final reports should include: - **Executive summary** (3–5 bullets, fits on one screen) - **Detailed findings** organized by theme, with citations - **Open questions** — what we still don't know - **Recommendations** — what to do based on the findings ## Asking Good Questions When I say "research X", before diving in, confirm: 1. What format do I want the output in? 2. Is this for a specific decision? If so, what is it? 3. How deep — quick scan (1–2 hours) or thorough (deep dive)?
Running prompts needs a free account.
Sign in and we'll stream the response from Claude Sonnet 4.6 right here — no config needed for the platform models.
Research Question [What are we trying to find out? Be specific — vague questions produce vague answers.]
# Research Project ## Research Question [What are we trying to find out? Be specific — vague questions produce vague answers.] ## Why This Matters [What decision or action does this research inform?] ## Scope & Constraints - Time horizon: [e.g. "focus on last 3 years"] - Geographic scope: [e.g. "US market only"] - Out of scope: [what we're explicitly NOT researching] ## File Structure ``` research/ ├── CLAUDE.md ← You are here ├── sources/ # Raw source material — one file per source │ └── TEMPLATE.md # Copy this for each new source ├── notes/ # Running notes, hypotheses, open questions │ ├── questions.md # Questions to answer (checked off as answered) │ └── findings.md # Key findings as they emerge ├── drafts/ # Work-in-progress documents └── output/ # Final deliverables (reports, memos, presentations) ``` ## Research Process 1. **Define** — clarify the question, agree on scope (this file) 2. **Gather** — collect sources into `sources/`, one file each with URL + summary 3. **Synthesize** — update `notes/findings.md` as patterns emerge 4. **Draft** — produce structured draft in `drafts/` 5. **Finalize** — clean up into `output/` deliverable ## Source Standards - Every claim needs a source — cite URL, date, and author where possible - Rate source reliability: Primary (official data) / Secondary (analysis) / Opinion - When sources conflict, flag it explicitly — don't pick a side silently - Prefer recent sources unless historical context is the point ## Output Format Final reports should include: - **Executive summary** (3–5 bullets, fits on one screen) - **Detailed findings** organized by theme, with citations - **Open questions** — what we still don't know - **Recommendations** — what to do based on the findings ## Asking Good Questions When I say "research X", before diving in, confirm: 1. What format do I want the output in? 2. Is this for a specific decision? If so, what is it? 3. How deep — quick scan (1–2 hours) or thorough (deep dive)?