Research Tools Marketplace
Economics research skills for Claude Code — streamline R analytics workflows and manuscript preparation with integrated tools for data analysis, bibliography management, and publication formatting.
📦 What's Included
This marketplace provides three integrated plugin collections with 12+ skills, MCP servers, and editor intelligence:
R workflow optimization for research data analysis
| Skill | Purpose |
|---|
| r-parallel | Convert sequential R code to parallel execution (future ecosystem) |
| r-style | Detect and refactor LLM code smells in R scripts |
| renv-manager | Set up and manage reproducible R environments |
| dependency-tracker | Map data flow and identify stale outputs in research pipelines |
Academic manuscript preparation with citation, formatting, and bibliography tools
MCP Servers
- semantic-scholar-mcp — Search Semantic Scholar API for papers and citations
- crossref-mcp — Query CrossRef metadata and publication versions
- bibtex-mcp — Local BibTeX parsing, citation scanning, and management
Skills
| Skill | Purpose |
|---|
| aer-figures | AER-compliant figure formatting and validation |
| aer-tables | AER-compliant table formatting and validation |
| bibtex-curator | BibTeX reference curation and quality control |
| bibtex-janitor | Automated BibTeX cleanup and normalization |
| notation-guardian | Mathematical notation consistency checking |
| palette-designer | Color palette design with auto-preview |
| proof-checker | Cross-reference and formatting proofreading |
| visual-sync | Visual consistency across manuscript elements |
Lightweight language-server support for R inside Claude Code. This gives Claude immediate feedback about your R files while it edits them, so it can catch errors earlier, navigate definitions, and format consistently. It uses Posit Air when available, and falls back to the languageserver R package if Air isn’t installed.
Python language-server support via ty (Astral). Gives Claude real-time type checking and diagnostics for .py and .pyi files while it edits them. The LSP also lets Claude navigate definitions and find references directly, rather than searching through files manually — reducing token usage in Python-heavy projects.
🚀 Installation
To install the plugins, first install the marketplace. Go inside Claude Code and run:
/plugin marketplace add halidaee/econtools_marketplace
Then, install the plugins:
/plugin install analytics-toolkit@econtools_marketplace
/plugin install manuscript-tools@econtools_marketplace
/plugin install r-lsp@econtools_marketplace
⚠️ Token Usage Note
These plugins include multiple MCP servers and skills that are loaded into Claude's context when active. While powerful, they can be token-heavy if all plugins remain enabled simultaneously.
Two approaches to manage token usage:
Option 1: Global install + enable/disable as needed
Install both plugins globally, then disable when not actively using them:
/plugin disable analytics-toolkit # When not working with R code
/plugin disable manuscript-tools # When not preparing manuscripts
Re-enable when needed:
/plugin enable analytics-toolkit # When working with R code
/plugin enable manuscript-tools # When preparing manuscripts
Option 2: Project-specific installation
Install plugins only in the projects where you need them. This prevents them from loading in unrelated work.
Even within a single research project, you may want to toggle between plugins depending on your task—analytics tools for data analysis, manuscript tools for writing. You won't typically need both active at the same time.
🔧 Configuration
CrossRef Email (Required for manuscript-tools)
The CrossRef MCP requires an email address to access their "polite pool" API tier (50 req/s vs 1 req/s). This takes 30 seconds to set up:
Option 1: Environment variable
export CROSSREF_MAILTO="your.email@domain.com"
Add this to your shell profile (~/.zshrc or ~/.bashrc) to make it permanent.
halidaee