Add Messy Column Fixer recipe

documentation/src/pages/recipes/data/recipes/messy-column-fixer.yaml

@@ -0,0 +1,43 @@
+version: 1.0.0
+title: Messy Column Fixer
+author:
+  contact: the-matrixneo 
+description: "Fixes messy columns: normalizes and cleans CSV data."
+instructions: |
+  1. Provide the path to your CSV file.
+  2. The recipe will scan all columns for type mismatches and missing values.
+  3. It will suggest fixes (or automatically apply them, depending on your choice).
+  4. Review the output and save your cleaned file.
+activities:
+  - Validate the input CSV file.
+  - Analyze columns for data quality issues (mixed types, missing values).
+  - Apply or suggest data cleaning and normalization fixes.
+  - Generate a summary report and the cleaned CSV file.
+  - Provide the cleaned CSV file with a "_cleaned" suffix.
+parameters:
+  - key: file_path
+    input_type: string
+    requirement: required
+    description: "Path to the CSV file you want to clean."
+  - key: auto_fix_decision
+    input_type: string
+    requirement: optional
+    description: "Describe how fixes should be applied (e.g., 'apply automatically', 'suggest only')."
+    default: "suggest only"
+    choices:
+      - "apply automatically"
+      - "suggest only"
+extensions:
+  - type: builtin
+    name: developer
+    description: "Fixes messy columns in CSV files by normalizing and cleaning the data."
+    display_name: Developer
+    timeout: 300
+    bundled: true
+prompt: |
+  You are a CSV cleaning assistant.
+  1.  First, validate that the file at {{ file_path }} exists and is a readable CSV. If not, inform the user and stop.
+  2.  Scan the file to identify columns with mixed data types, missing values, or formatting issues.
+  3.  Based on the {{ auto_fix_decision }} parameter, either suggest or apply fixes for the detected issues.
+  4.  For each fix, briefly explain the reasoning (e.g., "Converted 'Age' column to Integer because many values are numeric.").
+  5.  Provide a comprehensive summary of the changes and output the cleaned dataset.