# Research Idea Screen

> Use when a research idea or hypothesis must be screened against literature, current repo overlap, existing baselines/runs/specs, novelty risks, comparison papers, and minimum credible experiment direction before coding.

- Skill: `woosook0127/research-idea-screen` (Agent Skill)
- Install (CLI): `npx skillmds@latest add woosook0127/research-idea-screen`
- Raw SKILL.md: https://api.skillmd.com/api/skills/woosook0127/research-idea-screen/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: woosook0127 (https://skillmd.com/u/woosook0127)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/woosook0127/research-idea-screen

---


# Research Idea Screen

Use this before implementation when the user is testing a research direction.

## Procedure

1. Restate the idea as a falsifiable hypothesis.
2. Search local project knowledge first:
   - `research_knowledge/`
   - existing paper cards and Q&A
   - experiment specs, results, runs, model packages
3. If literature evidence is insufficient, invoke `literature-field-surveyor` or run a targeted paper search.
4. Identify the strongest comparison papers and baselines.
5. Map the idea to current repo capability with `framework-mapping` when needed.
6. Recommend: proceed, refine, defer, or drop.

## Output

- `IDEA`-style packet if durable state is needed
- repo overlap check
- literature signals and comparison papers
- minimum credible baseline comparison
- proposed next skill/team

## Guardrails

- Research is not a default coding team.
- Novelty claims require both literature and repo-overlap evidence.
- If evidence is weak, label the recommendation as provisional.