FAOSTAT Skills
AI-powered analysis skills for FAOSTAT, the world's leading food and agriculture statistics database.
FAOSTAT Skills
AI-powered analysis skills for the UN FAOSTAT database, the world's most comprehensive source of food and agriculture statistics. These platform-agnostic skills guide your AI assistant through multi-step analytical workflows: country food security profiles, trade dependency analysis, commodity deep dives, agrifood climate assessments, and data-driven storytelling.
Works with: Claude Code, OpenAI Codex, and any AI assistant that supports the SKILL.md format.
Quick Start
Claude Code (CLI)
# 1. Add this repo as a marketplace from CC CLI (one-time)
claude
/plugin marketplace add berba-q/faostat-skills
# 2. Install
/plugin install faostat-skills@faostat
# 3. Activate
/reload-plugins
OpenAI Codex
# Clone, then copy each skill folder directly into your skills directory
git clone https://github.com/berba-q/faostat-skills.git /tmp/faostat-skills
cp -r /tmp/faostat-skills/skills/* ~/.codex/skills/
Other AI assistants — copy the contents of the skills/ directory to wherever your tool discovers skill files. Consult your provider's documentation for the correct skills directory path.
Once installed, try asking:
"Give me a food security profile for Kenya"
"Compare wheat yields in France, USA, and India"
"How dependent is Egypt on wheat imports?"
Prerequisites
The FAOSTAT MCP Server must be installed and configured before using these skills.
pip install faostat-mcp
Then configure your credentials (one-time setup) by asking your AI assistant:
"Set up FAOSTAT with my credentials"
Installation
Claude Code
Claude Code uses a marketplace model. Add this repo as a marketplace source once, then install:
# Step 1 — add marketplace (one-time setup)
/plugin marketplace add berba-q/faostat-skills
# Step 2 — install the plugin (inside Claude Code)
/plugin install faostat-skills@faostat
Updates are automatic — Claude Code checks for new versions at startup and updates in the background. Run /reload-plugins to activate the latest version in your current session.
OpenAI Codex
Each skill folder must sit directly inside your Codex skills directory — not nested inside a parent folder. Clone and copy:
git clone https://github.com/berba-q/faostat-skills.git /tmp/faostat-skills
cp -r /tmp/faostat-skills/skills/* ~/.codex/skills/
Or keep the repo for easy updates via git pull:
git clone https://github.com/berba-q/faostat-skills.git ~/faostat-skills
for dir in ~/faostat-skills/skills/*/; do
ln -s "$dir" ~/.codex/skills/"$(basename "$dir")"
done
Consult the Codex documentation for the exact skills directory path on your system.
Other AI Assistants
Copy the skills/ directory to wherever your AI tool discovers skill files. Each skill is a self-contained SKILL.md with YAML frontmatter and markdown instructions.
Available Skills
Tier 1: Core Analysis
| Skill | Command | Description |
|---|---|---|
| Country Food Security Profile | /faostat-country-profile | Comprehensive food security assessment — production, trade, nutrition, and risk indicators for any country |
| Comparative Agricultural Analysis | /faostat-compare | Side-by-side comparison of agricultural metrics for two or more specific named entities |
| Commodity Deep Dive | /faostat-commodity | Complete global briefing for any commodity — production rankings, yield trends, trade flows |
| Trade Dependency Analyzer | /faostat-trade | Import dependence assessment with self-sufficiency ratios and supply chain risk indicators |
| Agrifood Climate Analyzer | /faostat-climate | Climate-agriculture nexus analysis — emissions profiles, temperature trends, land use, inputs-emissions links |
| Agricultural Trend Monitor | /faostat-trends | Identify biggest changes and anomalies in agricultural data over a time window |
Tier 2: Outputs
| Skill | Command | Description |
|---|---|---|
| Choropleth Map | /faostat-map | Interactive world map of any FAOSTAT country-level metric |
| Data Visualizer | /faostat-viz | Interactive Chart.js HTML charts — use when the chart itself is the deliverable |
| Infographic | /faostat-infographic | Shareable single-page visual summary for social, press, or pitch decks |
| Data Storyteller | /faostat-story | Data-driven narrative article with embedded charts for research |
| Analytical Brief | /faostat-analytical-brief | Multi-page policymaker-facing brief in FAOSTAT house style (PDF + xlsx appendix) |
| Scientific Paper | /faostat-scientific-paper | Peer-reviewable IMRaD research paper (docx + xlsx + bib) |
Tier 3: Utilities
| Skill | Command | Description |
|---|---|---|
| Data Export | /faostat-export-dataset | Clean tabular data bundle (xlsx + csv + data dictionary) — when data itself is the deliverable |
| FAOSTAT Explorer | /faostat-explore | Guided discovery of FAOSTAT's data catalog with sample data and explanations |
Examples
> Give me a food security profile for Kenya
> Compare wheat yields in France, USA, and India over the last 10 years
> Tell me everything about global rice production and trade
> What are the agrifood emissions for Brazil? How do they compare per capita?
> How dependent is Egypt on wheat imports?
> Chart the top 10 maize producers over the last 20 years
> What's changed most in African agriculture over the last 5 years?
> Help me write a data story about the global avocado boom
> What data does FAOSTAT have about fertilizers?
FAOSTAT Domain Reference
Key domains used by these skills:
| Domain | Content |
|---|---|
| QCL | Crops and Livestock Products (production, area, yield) |
| TM | Trade Matrix (import/export by partner) |
| FBS | Food Balance Sheets (supply/demand for 360+ foods) |
| FS | Food Security indicators |
| GT | Emissions Totals — agrifood systems (broadest scope) |
| GCE | Emissions from crops only |
| GF | Forest emissions/removals — sinks, net conversion |
| EI | Emissions Intensities |
| EM | Emissions Indicators (shares, per capita, per value) |
| ET | Temperature Change |
| RL | Land Use |
| RFN/RFM/RFB | Fertilizers |
| RP/RT | Pesticides use and trade |
Architecture
FAOSTAT Skills (this repo) FAOSTAT MCP Server (separate)
┌────────────────────────┐ ┌────────────────────────┐
│ 14 platform-agnostic │ uses │ 21 MCP tools │
│ SKILL.md workflows │────────→ │ Published on PyPI │
│ │ │ faostat-mcp · stable │
└────────────────────────┘ └────────────────────────┘
Release flow:
npm version patch # bumps all 4 version fields atomically (see below)
git push && git push --tags
npm publish
Version bumping is handled by scripts/sync-version.mjs, wired into npm's version lifecycle hook in package.json. When you run npm version patch, npm updates package.json, the hook propagates the same version to .claude-plugin/plugin.json, marketplace.json, and .claude-plugin/marketplace.json, then stages all four for the version commit. This prevents the drift bug where .claude-plugin/plugin.json silently stayed at 0.1.0 across four published npm releases — Claude Code reads its displayed plugin version from that file, not from package.json.
Skills orchestrate the MCP server's tools into multi-step analysis workflows. They encode FAOSTAT domain expertise so users can easily interact with the FAOSTAT data.
Project Structure
faostat-skills/
├── skills/ ← Core: 14 platform-agnostic SKILL.md files
│ ├── country-profile/ ← Tier 1: analysis
│ ├── compare/
│ ├── commodity/
│ ├── trade/
│ ├── trends/
│ ├── climate/
│ ├── explore/
│ ├── viz/ ← Tier 2: outputs
│ ├── map/
│ ├── infographic/
│ ├── story/
│ ├── analytical-brief/
│ ├── scientific-paper/
│ └── export-dataset/ ← Tier 3: utilities
├── .claude-plugin/ ← Claude Code packaging
├── commands/ ← Claude Code slash commands
├── README.md
└── LICENSE
Contributing
- Fork the repository
- Create a feature branch (
git checkout -b feature/new-skill) - Follow the existing SKILL.md format (YAML frontmatter + markdown instructions)
- Test the skill with a real FAOSTAT MCP server connection
- Submit a pull request
License
MIT License - see LICENSE for details.
Related
- FAOSTAT MCP Server — The MCP server these skills are built on
- FAOSTAT — The UN FAO FAOSTAT database