{"id":19211,"plugin_id":"plugins_6a92f40359bc81919e5b41b1451ec501","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T23:15:23.649Z","digest":"7bba3ede849c5d28ee7d6096717a59f78df99d4e048e58a62125c62ede6f84e6","against":null,"payload":{"description":"Use when adding causal and counterfactual thinking to Power BI sales forecasts, including working days, holidays, delivery constraints, price changes, product lifecycle, stockouts, campaigns, customer behavior, and best/base/worst case simulations.","included_files":[{"relative_path":"agents/openai.yaml","size_in_bytes":144}],"name":"powerbi-causal-counterfactual-forecasting","skill_md_contents":"---\r\nname: powerbi-causal-counterfactual-forecasting\r\ndescription: Use when adding causal and counterfactual thinking to Power BI sales forecasts, including working days, holidays, delivery constraints, price changes, product lifecycle, stockouts, campaigns, customer behavior, and best/base/worst case simulations.\r\n---\r\n\r\n# Power BI Causal Counterfactual Forecasting\r\n\r\nUse this skill when the user asks why a forecast changes, what causes a revenue gap, or what would happen under alternative assumptions.\r\n\r\n## Feature families\r\n\r\n- calendar: working days, holidays, month length, fiscal periods\r\n- order flow: order age, requested delivery, planned delivery, status, backlog value\r\n- customer behavior: recency, frequency, average order value, churn or reactivation signals\r\n- product lifecycle: new product, mature product, discontinued product, replacement product\r\n- operations: supply constraints, delivery delay, stockout indicators\r\n- commercial: price change, discounting, campaign, sales initiative, budget/roll assumptions\r\n\r\n## Workflow\r\n\r\n1. Separate correlation from actionable cause. Do not claim causality without a plausible mechanism and supporting time sequence.\r\n2. Build counterfactuals:\r\n   - if backlog conversion improves\r\n   - if delivery slips\r\n   - if customer demand follows prior year\r\n   - if budget pressure is ignored\r\n   - if low-confidence segments are excluded\r\n3. Quantify sensitivity:\r\n   - revenue impact\r\n   - probability\r\n   - confidence\r\n   - affected customer/product/month\r\n4. Explain the causal story in one sentence per material driver.\r\n\r\n## Required outputs\r\n\r\n- `forecast_month`\r\n- `driver`\r\n- `driver_type`\r\n- `base_value`\r\n- `counterfactual_value`\r\n- `revenue_impact`\r\n- `confidence`\r\n- `evidence`\r\n- `actionability`\r\n\r\n## Guardrails\r\n\r\n- Mark drivers as `hypothesis` when the data only supports association.\r\n- Avoid overfitting small customer/product segments.\r\n- Use backtests to prove that adding a driver improves WAPE or bias before making it a default weight.\r\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}