feat: remove download_and_parse_resource tool (#66)

The tool caused RAM and disk storage issues on production MCP server
instances due to in-memory file parsing and potential large downloads.

The Tabular API (query_resource_data) covers the main use cases, and
LLMs can fall back to fetching raw file URLs directly via get_resource_info
for unsupported formats or files exceeding Tabular API limits.

Made-with: Cursor

Author: bolinocroustibat

README.md

@@ -344,13 +344,7 @@ The MCP server provides tools to interact with data.gouv.fr datasets and dataser
 
   Parameters: `question` (required), `resource_id` (required), `page` (optional, default: 1), `page_size` (optional, default: 20, max: 200)
 
-  Note: Recommended workflow: 1) Use `search_datasets` to find the dataset, 2) Use `list_dataset_resources` to see available resources, 3) Use `query_resource_data` with default `page_size` (20) to preview data structure. For small datasets (<500 rows), increase `page_size` or paginate. For large datasets (>1000 rows), use `download_and_parse_resource` instead. Works for CSV/XLS resources within Tabular API size limits (CSV ≤ 100 MB, XLSX ≤ 12.5 MB).
-
-- **`download_and_parse_resource`** - Download and parse a resource that is not accessible via Tabular API (files too large, formats not supported, external URLs).
-
-  Parameters: `resource_id` (required), `max_rows` (optional, default: 20), `max_size_mb` (optional, default: 500)
-
-  Supported formats: CSV, CSV.GZ, JSON, JSONL. Useful for files exceeding Tabular API limits or formats not supported by Tabular API. Start with default max_rows (20) to preview, then call again with higher max_rows if you need all data.
+  Note: Recommended workflow: 1) Use `search_datasets` to find the dataset, 2) Use `list_dataset_resources` to see available resources, 3) Use `query_resource_data` with default `page_size` (20) to preview data structure. For small datasets (<500 rows), increase `page_size` or paginate. For large datasets (>1000 rows), continue paginating or use `get_resource_info` to retrieve the raw file URL and fetch it directly. Works for CSV/XLS resources within Tabular API size limits (CSV ≤ 100 MB, XLSX ≤ 12.5 MB).
 
 ### Dataservices (external APIs)
 

tools/__init__.py

@@ -1,8 +1,5 @@
 from mcp.server.fastmcp import FastMCP
 
-from tools.download_and_parse_resource import (
-    register_download_and_parse_resource_tool,
-)
 from tools.get_dataservice_info import register_get_dataservice_info_tool
 from tools.get_dataservice_openapi_spec import (
     register_get_dataservice_openapi_spec_tool,
@@ -26,5 +23,4 @@ def register_tools(mcp: FastMCP) -> None:
     register_get_dataset_info_tool(mcp)
     register_list_dataset_resources_tool(mcp)
     register_get_resource_info_tool(mcp)
-    register_download_and_parse_resource_tool(mcp)
     register_get_metrics_tool(mcp)

tools/download_and_parse_resource.py

@@ -11,8 +11,8 @@ async def get_resource_info(resource_id: str) -> str:
         Get detailed information about a specific resource (file).
 
         Returns format, size, MIME type, URL, and checks Tabular API availability.
-        Helps decide which tool to use: query_resource_data (if Tabular API available)
-        or download_and_parse_resource (for large files or unsupported formats).
+        Helps decide whether to use query_resource_data (if Tabular API is available)
+        or fetch the raw file URL directly for unsupported formats or large files.
         """
         try:
             # Get full resource data from API v2

tools/get_resource_info.py

@@ -10,8 +10,8 @@ async def list_dataset_resources(dataset_id: str) -> str:
         List all resources (files) in a dataset with their metadata.
 
         Returns resource ID, title, format, size, and URL for each file.
-        Next step: use query_resource_data for CSV/XLSX files,
-        or download_and_parse_resource for other formats (JSON, JSONL) or large datasets.
+        Next step: use query_resource_data for CSV/XLSX files via the Tabular API,
+        or fetch the resource URL directly for other formats (JSON, JSONL) or large datasets.
         """
         try:
             dataset = await datagouv_api_client.get_dataset_details(dataset_id)