mcp-french-associations
# mcp-french-associations
MCP server for French associations: RNA, JOAFE, grants/subventions, and public dataset discovery.
## Tools
Run the MCP and call `french_associations_get_sources` first to inspect source coverage. This server also exposes domain-specific tools for the topic described above.
- `french_associations_search_datasets`
- `french_associations_get_dataset`
- `french_associations_build_search_plan`
- `french_associations_search_subsidies`
## Install
```bash
npm install
npm run build
npm test
npm run dev
```
## Claude Desktop
```json
{
"mcpServers": {
"french-associations": {
"command": "npx",
"args": ["mcp-french-associations"]
}
}
}
```
## Sources
- RNA data on data.gouv.fr: https://www.data.gouv.fr/datasets/repertoire-national-des-associations/
- JOAFE: https://www.journal-officiel.gouv.fr/pages/associations/
- Public subsidies data: https://www.data.gouv.fr/datasets/?q=subventions+associations
- data.gouv.fr API: https://doc.data.gouv.fr/api/reference/
## Publishing
See [docs/publishing.md](docs/publishing.md).
## Glama / Docker
The repo includes `Dockerfile` and `glama.json`.
Build steps:
```json
["npm install", "npm run build"]
```
CMD arguments:
```json
["node", "dist/index.js"]
```
## Safety
This MCP helps agents discover and summarize public sources. It is not an official authority. Verify decisions against the competent public service or original data producer.
## License
MIT
TDQS
Scored across 6 tools
Each tool has a clear primary purpose, though search_datasets and search_subsidies overlap in scope (both search data.gouv.fr for association-related datasets). The descriptions specify search_datasets covers RNA, JOAFE, subsidies, and grants, while search_subsidies is narrower, so they are distinguishable but could cause minor misselection.
All tool names follow a consistent pattern: the prefix 'french_associations_' combined with a verb_noun structure (get_sources, fetch_source_excerpt, search_datasets, get_dataset, build_search_plan, search_subsidies). This makes the tool set predictable and easy to navigate.
With only 6 tools, the server is highly focused and each tool contributes to its stated purpose of finding and inspecting French association data. The count is well-scoped, avoiding unnecessary bloat or overly sparse coverage.
The tools cover listing sources, fetching excerpts, searching for datasets, inspecting datasets, and building search plans. A notable gap is the absence of a tool to actually retrieve full dataset records or query resources directly, but the server appears designed to assist agents in planning searches rather than executing heavy data pulls, so this is a minor gap.