31 MCP tools for ESG data extraction, PDF processing, vector search, and EU regulation analysis.
Open-source Model Context Protocol servers for ESG (Environmental, Social, and Governance) data extraction, analysis, and regulation management.
31 tools across 6 servers — install once, run only what you need.
Author: Ioannis Michos (johnmichos.tf@gmail.com)
pip install esg-mcp-servers
| Service | Required for |
|---|---|
| PostgreSQL 16 + pgvector | Vector storage, metrics, regulations |
| MongoDB 7 | PDF binary storage (GridFS) |
| Anthropic API key | RAG queries, metric extraction |
Spin up local databases with Docker:
docker compose up -d
Run the database migration:
esg-mcp-migrate
Copy .env.example and fill in your values:
cp .env.example .env
Key variables:
| Variable | Default | Description |
|---|---|---|
POSTGRES_DSN | postgresql://esg:esg@localhost/esg_platform | PostgreSQL connection string |
MONGODB_URI | mongodb://localhost:27017 | MongoDB connection string |
ANTHROPIC_API_KEY | — | Required for RAG and LLM extraction |
EMBEDDING_MODEL | Snowflake/snowflake-arctic-embed-l-v2.0 | Sentence-transformer model |
EMBEDDING_DIMENSIONS | 1024 | Embedding vector size |
ESRS-aligned KPI extraction for emissions, energy, water, waste, social, and governance domains.
| Tool | Description |
|---|---|
extract_emissions_data | GHG Scope 1, 2, 3 extraction |
extract_energy_data | Energy consumption & renewables (ESRS E2) |
extract_water_data | Water withdrawal, discharge, consumption (ESRS E3) |
extract_waste_data | Waste generation, recycling, landfill (ESRS E5) |
extract_social_data | Workforce, diversity, H&S (ESRS S1) |
extract_governance_data | Board composition & governance (ESRS G1) |
answer_esg_query | Free-text RAG Q&A over documents |
keyword_similarity_search | Semantic keyword search in documents |
detect_query_domain | Classify query into ESG domain |
detect_emissions_query_type | Detect emissions scope/year from query |
batch_extract_metrics | Extract all KPIs and persist to DB |
PDF validation, text/table extraction, and embedding generation.
| Tool | Description |
|---|---|
verify_esg_report | RandomForest ESG report classifier |
extract_text_chunks | Extract and chunk PDF text |
extract_tables | Table extraction with OCR fallback |
generate_embeddings | Batch embedding generation |
process_pdf_full_pipeline | End-to-end: extract, embed, store |
pgvector CRUD operations for document chunks and query cache.
| Tool | Description |
|---|---|
upsert_document_chunks | Insert/update chunks with embeddings |
similarity_search | Cosine similarity search |
get_cached_query_response | Retrieve cached LLM responses |
cache_query_response | Store LLM response in cache |
list_documents | List indexed documents |
EU ESG regulation download, ingestion, and semantic search.
| Tool | Description |
|---|---|
download_regulation | Download regulation PDF from EUR-Lex |
download_all_regulations | Batch download all configured regulations |
ingest_regulation | Extract, parse articles, embed, store |
search_regulation_text | Semantic search across regulation articles |
list_regulations | List ingested regulations |
ESG report discovery and download from the web.
| Tool | Description |
|---|---|
search_esg_reports | Multi-engine search for ESG PDFs |
crawl_company_website | Deep-crawl website for PDF links |
download_pdf | Download PDF and store in GridFS |
get_scraping_status | Check scrape job status |
cancel_scraping_job | Cancel running scrape job |
All tools from all servers in a single process.
Add to your claude_desktop_config.json:
{
"mcpServers": {
"esg-metrics": {
"command": "esg-metrics-extractor"
},
"esg-pdf": {
"command": "esg-pdf-processor"
},
"esg-vectors": {
"command": "esg-vector-store"
},
"esg-regulations": {
"command": "esg-regulations"
},
"esg-scraper": {
"command": "esg-scraper"
}
}
}
Or use the combined server for all tools:
{
"mcpServers": {
"esg": {
"command": "esg-mcp-all"
}
}
}
The scraper server requires Selenium for JavaScript-heavy sites:
pip install esg-mcp-servers[scraper]
git clone https://github.com/freminder/esg-mcp-servers.git
cd esg-mcp-servers
pip install -e ".[scraper,dev]"
docker compose up -d
esg-mcp-migrate
MIT — see LICENSE.
Source-derived launch command. Check the maintainer’s required arguments and credentials before running:
uvx esg-mcp-serversMerge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.
{
"mcpServers": {
"io-github-freminder-esg-mcp-servers": {
"command": "uvx",
"args": [
"esg-mcp-servers"
]
}
}
}Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.
Claude Desktop setup referenceesg-mcp-serverspypiESG MCP Servers works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.
~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect.~/.cursor/mcp.jsonRestart Cursor for changes to take effect..vscode/mcp.jsonReload VS Code window for changes to take effect.~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect..mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.