agentleFS
Felpix-StudiosFelpix-Studios/social-science-research/skills/research-setup/SKILL.md

research-setup

Skill8 starsChanged 4 months agoMITView on GitHub

Interactive wizard that configures the social-science-research plugin for a new project. Asks about field, institution, journals, datasets, key researchers, R palette; writes answers to references/domain-profile.md and CLAUDE.md. Use first when starting fresh or configuring the plugin.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 6 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read ${CLAUDE_PLUGIN_ROOT}/skills/research-setup/references/workflow-overview.md before explaining the workflow or next-step sequence.

    line 15
  2. Very specific:

    Seed references/domain-profile.md if absent: Run via Bash (idempotent — only copies when missing):

    line 22
  3. Distinctive:

    Seed CLAUDE.md if absent: Use Glob to check. If missing, Write this placeholder:

    line 31
  4. Distinctive:

    Ask conversationally (not via AskUserQuestion):

    line 54
  5. Distinctive:

    Wait for the user's response before moving to Step 3.

    line 60
  6. Distinctive:

    If the user chose "Add researchers", ask conversationally for names, institutions, and Google Scholar URLs.

    line 91
  7. Distinctive:

    If the user chose "Custom colors", ask conversationally for primary and secondary hex codes.

    line 93
  8. Very specific:

    IMPORTANT: Do NOT write to ${CLAUDE_PLUGIN_ROOT}/rules/r-code-conventions.md.

    line 102
  9. Distinctive:

    Update the project header placeholders:

    line 106

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Research Setup Wizard

Walk the user through setup using AskUserQuestion menus wherever possible. Collect all answers, then write config files in one pass.

This skill is the only entry point that seeds project config (references/domain-profile.md, CLAUDE.md). The plugin does not scaffold anything on session start.

Read ${CLAUDE_PLUGIN_ROOT}/skills/research-setup/references/workflow-overview.md before explaining the workflow or next-step sequence.

Step 0: Bootstrap & Check Existing Config

Seed missing config files into the project, then read what's there.

  1. Seed references/domain-profile.md if absent. Run via Bash (idempotent — only copies when missing):

    if [ ! -f ./references/domain-profile.md ]; then
      mkdir -p ./references
      cp "${CLAUDE_PLUGIN_ROOT}/references/domain-profile.md" ./references/domain-profile.md
    fi
  2. Seed CLAUDE.md if absent. Use Glob to check. If missing, Write this placeholder:

    # [YOUR PROJECT NAME]
    
    **Author:** [YOUR NAME]
    **Institution:** [YOUR INSTITUTION]
  3. Now Read CLAUDE.md and references/domain-profile.md silently. If either has real content (not placeholders), tell the user what's already configured and offer to confirm or update each section.

Step 1: Field

AskUserQuestion — 1 question:

  1. Field (single select, header: "Field"): "What is your field or subfield?" — Options: "Political Science", "Economics", "Sociology", "Public Health". User selects Other for unlisted fields.

Store field for customizing later steps.

Step 2: Identity & Project Title

Ask conversationally (not via AskUserQuestion):

  1. "What is your name (for paper authorship)? You can skip this for now."
  2. "What is your institution? (This helps customize figure color palettes and LaTeX title pages.)"
  3. "Do you have a working title for your current project? Fine to skip if you're still exploring."

Wait for the user's response before moving to Step 3.

Step 3: Journals

AskUserQuestion — 2 multi-select questions based on the user's field:

  1. Top journals (multiSelect, header: "Top journals"): "Select your top journals (pick all that apply):" — Offer the 4 most prestigious journals for the user's selected field. Use your knowledge of the field's top venues.
  2. Secondary journals (multiSelect, header: "Secondary"): "Select secondary or subfield journals:" — Offer 4 secondary/subfield journals appropriate to the user's field.

User can multi-select presets and add custom journals via Other.

Step 4: Datasets

AskUserQuestion — 2 multi-select questions based on the user's field:

  1. Common datasets (multiSelect, header: "Datasets"): "Which datasets do you commonly use?" — Offer the 4 most-used datasets for the user's field. Use your knowledge of field-standard data sources.
  2. Additional sources (multiSelect, header: "More data"): "Any additional data sources?" — Offer 4 cross-disciplinary sources: World Bank Open Data, IPUMS, Bureau of Labor Statistics, Census Bureau.

Step 5: Researchers & Colors

AskUserQuestion — 2 questions at once:

  1. Researchers (single select, header: "Researchers"): "Add key researchers for targeted literature searches?" — Options: "Skip for now" / "Add researchers".
  2. Colors (single select, header: "Colors"): "Which color palette should R figures use?" — If the institution is well-known, offer its official brand colors as the recommended option with a preview showing the hex codes. Otherwise offer: "Custom colors" / "Skip for now".

If the user chose "Add researchers", ask conversationally for names, institutions, and Google Scholar URLs. Do NOT guess researcher names.

If the user chose "Custom colors", ask conversationally for primary and secondary hex codes.

Step 6: Write Config Files

references/domain-profile.md

Read the existing file and update each section with the user's answers. Preserve the file's existing structure and comments. Only replace sections the user answered; leave unanswered sections unchanged. Add an "Institutional Colors" section at the end if one doesn't exist, with the user's hex codes (or defaults #012169 / #f2a900 if skipped).

IMPORTANT: Do NOT write to ${CLAUDE_PLUGIN_ROOT}/rules/r-code-conventions.md. Colors live in references/domain-profile.md only.

CLAUDE.md

Update the project header placeholders:

  • [YOUR PROJECT NAME] → user's project title (or leave placeholder if skipped)
  • [YOUR NAME] → user's name (or leave placeholder if skipped)
  • [YOUR INSTITUTION] → user's institution

Step 7: Confirm

Setup complete:
- references/domain-profile.md — field, [N] journals, [N] datasets[, N researchers][, colors]
- CLAUDE.md — project name and institution

Update anytime: references/domain-profile.md
Next step: /new-project

Note any skipped sections and where to add them later.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.