Update to SQL Copilot

{
  "description": "Write ranking, running totals, top-N and deterministic deduplication SQL.",
  "included_files": [],
  "name": "window-ranking-dedup",
  "skill_md_contents": "---\nname: window-ranking-dedup\ndescription: Write ranking, running totals, top-N and deterministic deduplication SQL.\n---\n\n# Window Ranking Dedup\n\n**Trigger:** Write ranking, running totals, top-N and deterministic deduplication SQL.\n\n## Workflow\n1. Define the partition key, desired row grain, order direction and precise tie behavior.\n2. Choose ROW_NUMBER, RANK or DENSE_RANK according to ties, or an aggregate/window when row retention differs.\n3. For latest-record selection, add a complete deterministic order using version, updated timestamp and final immutable unique tie-breaker.\n4. Filter window results at the correct logical level using a CTE/subquery or dialect-supported QUALIFY.\n5. Include validation that one winner is selected per business key and explain whether a re-run with late data changes the winner.\n\n## Critical standard\nNo nondeterministic ROW_NUMBER with an incomplete ORDER BY for production deduplication.\n\n## Evidence and execution\nOnly say \"executed\", \"passed\", \"observed\", \"optimized\" or \"deployed\" when a connected tool returned evidence for that claim. An SQL draft, a textual review, or local fixture proof alone is not production validation. Never ask the user to paste passwords or raw confidential records.\n\n## Supporting references\n- Read `../../references/windows-and-dedup.md` when relevant.\n- Read `../../references/dialects.md` when relevant.\n\n## Return\nDialect-specific SQL with explicit tie semantics and uniqueness probes.\n"
}