What problem does it solve? Borough names appear inconsistently across Montréal open datasets — different dash styles, abbreviations, numeric codes, and informal neighborhood names — making geographic filtering error-prone. This Skill gives agents a canonical reference for resolving any borough mention into the correct dataset values. ## Core Features & Use Cases - Canonical Borough Reference: Lists all 19 boroughs with codes (AHU, CDN, VIM), common short names, and approximate center coordinates, plus the 15 reconstituted municipalities. - Name Variation Handling: Documents how borough names appear in dataset metadata (territoire codes), record fields (ARROND_NOM), and numeric fields, with robust SQL LIKE patterns for filtering. - Natural Language Mapping: Maps informal terms like "downtown", "Mile End", "HoMa", or "the West Island" to the correct official borough. - Use Case: A user asks "How many trees are in the Plateau?" — the agent resolves "the Plateau" to Le Plateau-Mont-Royal and builds a filter like WHERE "ARROND_NOM" LIKE '%Plateau%Royal%' that survives formatting inconsistencies. ## Quick Start Ask the agent to filter a Montréal dataset by a borough or neighborhood name, such as "show building permits in Verdun this month".
