# llm-research-workflow

llm-research-workflow enforces a structured methodology for extracting, classifying, and validating empirical variables in economics and finance research with LLMs.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf (Devin Desktop), Claude.ai · ★ 240 repo stars
- Author: franklee16
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/franklee16/academic-research-skills/tree/HEAD/research-workflow/llm-research-workflow/SKILL.md
- Updated: 2026-09-18
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/llm-research-workflow

## What it does

The skill integrates an academic methodology drawn from Ferrara (2026, NBER WP 35374) directly into Claude Code sessions. Instead of immediately writing automation scripts, the assistant pauses to evaluate three mandatory prerequisites: tool suitability, data transfer legality, and execution costs. Deterministic lookups or spatial queries get routed toward dedicated libraries and authoritative sources rather than raw generation.

The core loop requires offering multiple alternatives for every variable extraction task, explicitly including at least one deterministic or non-LLM baseline such as keyword rules or supervised algorithms. Claude guides the discussion through method tradeoffs, user comprehension, and mandatory pilot testing on representative subsamples before running full-scale pipelines.

Licensed academic feeds like CRSP, WRDS, or broker archives often forbid bulk uploads to external services. The skill halts blind execution and forces you to consider zero-retention contracts or self-hosted weights like Qwen or Llama.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add franklee16/academic-research-skills --skill llm-research-workflow -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/llm-research-workflow/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/llm-research-workflow/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add franklee16/academic-research-skills --skill llm-research-workflow -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/llm-research-workflow/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add franklee16/academic-research-skills --skill llm-research-workflow -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/llm-research-workflow/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install franklee16/academic-research-skills research-workflow/llm-research-workflow/SKILL.md
```

2. **Manually: put the files in place**
- File: `.github/skills/llm-research-workflow/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add franklee16/academic-research-skills --skill llm-research-workflow -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/franklee16/academic-research-skills --path research-workflow/llm-research-workflow
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/llm-research-workflow/SKILL.md` — The legacy .windsurf/skills/ path still works

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/franklee16/academic-research-skills/tree/HEAD/research-workflow/llm-research-workflow/SKILL.md

## When to use it
- verify whether commercial dataset licenses permit third-party API processing
- benchmark generative classification against deterministic non-LLM methods
- structure pilot testing on representative corpora prior to large-scale runs
- route deterministic queries away from prone-to-hallucination model generation

## Example prompts
- Can an LLM reliably extract balance sheet indicators from these PDF filings?
- Help me set up a pipeline to score sentiment in earnings conference calls
- Classify these historical regulatory notices for an empirical regressions study

## Similar
- [research-ideation](https://affpapa.org/en/nevibecode/skills/awesome-econ-ai-community-research-ideation.md) — skill · 196 installs
- [research-junshi](https://affpapa.org/en/nevibecode/skills/research-junshi.md) — skill · ★ 135 repo stars
- [k2f-research-deck](https://affpapa.org/en/nevibecode/skills/k2f-research-deck.md) — skill · 664 installs
- [pyeconomics](https://affpapa.org/en/nevibecode/mcp/pyeconomics.md) — MCP server · ★ 11 repo stars
- [founder-cfo](https://affpapa.org/en/nevibecode/skills/founder-cfo.md) — skill · 736 installs
- [AgentEcon](https://affpapa.org/en/nevibecode/mcp/agentecon.md) — MCP server · 3813 installs

From the same set: [franklee16/academic-research-skills](https://affpapa.org/en/nevibecode/by/franklee16/academic-research-skills.md) — 98 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
