miguel-escribano/openair-client

This skill should be used when loading CSV or Excel from the MCP server disk, or ADMS/AURN CSV files already on the server host. Triggers: csv, xlsx, disk path, load_series_from_csv, load_series_from_excel, import_adms, import_aurn_csv. Not for files on the user's PC, base64 upload, live public network import, or regional portal Excel (use ingest-local-export, remote-file-upload, ingest-network, or regional-excel).

This skill should be used when a CSV or Excel file is on the user's machine and the agent can run export_local_series.py before calling the remote openair MCP. Triggers: export script, local file remote MCP, SeriesV1 json, Windows excel path. Not for server disk paths, base64 upload, or live public network import.

This skill should be used when importing air quality data from live public monitoring networks via openair MCP. Triggers: import AURN, import europe, import ukaq, public network, monitoring station, site code, MY1. Not for local CSV/Excel files on disk.

This skill should be used when chaining another MCP's time-series JSON into openair plots via json_exports. Triggers: another MCP, API, postgres, airtable, json_exports, sensor platform, upstream MCP. Not for CSV/Excel on disk or public network import_*.

This skill should be used to route openair MCP ingest and plot requests to the correct specialist skill. Triggers: plot, chart, csv, excel, AURN, import europe, upload, json_exports, wind rose, calendar plot. Not for methodology questions (use openair_docs on the server).

This skill should be used when SeriesV1 data is already loaded and the user wants an openair chart via MCP. Triggers: plot, chart, time_plot, calendar_plot, time_variation, cor_plot, polar_plot, wind_rose, diurnal, trend. Not for raw file ingest (route to ingest-local, ingest-network, or multi-mcp first).