with last commit, we can simplify the logic + remove old 2025_ prefix for search of the tables

common/build.rs

@@ -29,6 +29,30 @@ fn main() -> Result<(), Box<dyn std::error::Error>> {
             ".komp_ac.table_validation.CountMode",
             "#[derive(serde::Serialize, serde::Deserialize)] #[serde(rename_all = \"SCREAMING_SNAKE_CASE\")]",
         )
+        .type_attribute(
+            ".komp_ac.table_definition.ColumnDefinition",
+            "#[derive(serde::Serialize, serde::Deserialize)]",
+        )
+        .type_attribute(
+            ".komp_ac.table_definition.TableLink",
+            "#[derive(serde::Serialize, serde::Deserialize)]"
+        )
+        .type_attribute(
+            ".komp_ac.table_definition.PostTableDefinitionRequest",
+            "#[derive(serde::Serialize, serde::Deserialize)]",
+        )
+        .type_attribute(
+            ".komp_ac.table_definition.TableDefinitionResponse",
+            "#[derive(serde::Serialize, serde::Deserialize)]"
+        )
+        .type_attribute(
+            ".komp_ac.table_script.PostTableScriptRequest",
+            "#[derive(serde::Serialize, serde::Deserialize)]",
+        )
+        .type_attribute(
+            ".komp_ac.table_script.TableScriptResponse",
+            "#[derive(serde::Serialize, serde::Deserialize)]",
+        )
         .compile_protos(
             &[
                 "proto/common.proto",

common/src/proto/komp_ac.table_definition.rs

@@ -1,4 +1,5 @@
 // This file is @generated by prost-build.
+#[derive(serde::Serialize, serde::Deserialize)]
 #[derive(Clone, PartialEq, ::prost::Message)]
 pub struct TableLink {
     #[prost(string, tag = "1")]
@@ -6,6 +7,7 @@ pub struct TableLink {
     #[prost(bool, tag = "2")]
     pub required: bool,
 }
+#[derive(serde::Serialize, serde::Deserialize)]
 #[derive(Clone, PartialEq, ::prost::Message)]
 pub struct PostTableDefinitionRequest {
     #[prost(string, tag = "1")]
@@ -19,6 +21,7 @@ pub struct PostTableDefinitionRequest {
     #[prost(string, tag = "5")]
     pub profile_name: ::prost::alloc::string::String,
 }
+#[derive(serde::Serialize, serde::Deserialize)]
 #[derive(Clone, PartialEq, ::prost::Message)]
 pub struct ColumnDefinition {
     #[prost(string, tag = "1")]
@@ -26,6 +29,7 @@ pub struct ColumnDefinition {
     #[prost(string, tag = "2")]
     pub field_type: ::prost::alloc::string::String,
 }
+#[derive(serde::Serialize, serde::Deserialize)]
 #[derive(Clone, PartialEq, ::prost::Message)]
 pub struct TableDefinitionResponse {
     #[prost(bool, tag = "1")]

common/src/proto/komp_ac.table_script.rs

@@ -1,4 +1,5 @@
 // This file is @generated by prost-build.
+#[derive(serde::Serialize, serde::Deserialize)]
 #[derive(Clone, PartialEq, ::prost::Message)]
 pub struct PostTableScriptRequest {
     #[prost(int64, tag = "1")]
@@ -10,6 +11,7 @@ pub struct PostTableScriptRequest {
     #[prost(string, tag = "4")]
     pub description: ::prost::alloc::string::String,
 }
+#[derive(serde::Serialize, serde::Deserialize)]
 #[derive(Clone, PartialEq, ::prost::Message)]
 pub struct TableScriptResponse {
     #[prost(int64, tag = "1")]

server/src/steel/server/execution.rs

@@ -3,21 +3,20 @@
 use steel::steel_vm::engine::Engine;
 use steel::steel_vm::register_fn::RegisterFn;
 use steel::rvals::SteelVal;
-use super::functions::{SteelContext, convert_row_data_for_steel};
+use super::functions::SteelContext;
 use steel_decimal::registry::FunctionRegistry;
 use sqlx::PgPool;
 use std::sync::Arc;
 use std::collections::HashMap;
 use thiserror::Error;
 use tracing::{debug, error};
+use regex::Regex; // NEW
 
-/// Represents different types of values that can be returned from Steel script execution.
 #[derive(Debug)]
 pub enum Value {
     Strings(Vec<String>),
 }
 
-/// Errors that can occur during Steel script execution.
 #[derive(Debug, Error)]
 pub enum ExecutionError {
     #[error("Script execution failed: {0}")]
@@ -28,7 +27,83 @@ pub enum ExecutionError {
     UnsupportedType(String),
 }
 
-/// Creates a Steel execution context with proper boolean value conversion.
+// NEW: upgrade steel_get_column -> steel_get_column_with_index using FK present in row_data
+fn auto_promote_with_index(
+    script: &str,
+    current_table: &str,
+    row_data: &HashMap<String, String>,
+) -> String {
+    // Matches: (steel_get_column "table" "column")
+    let re = Regex::new(
+        r#"\(\s*steel_get_column\s+"([^"]+)"\s+"([^"]+)"\s*\)"#,
+    )
+    .unwrap();
+
+    re.replace_all(script, |caps: &regex::Captures| {
+        let table = caps.get(1).unwrap().as_str();
+        let column = caps.get(2).unwrap().as_str();
+
+        // Only upgrade cross-table calls, if FK is present in the request data
+        if table != current_table {
+            let fk_key = format!("{}_id", table);
+            if let Some(id_str) = row_data.get(&fk_key) {
+                if let Ok(_) = id_str.parse::<i64>() {
+                    return format!(
+                        r#"(steel_get_column_with_index "{}" {} "{}")"#,
+                        table, id_str, column
+                    );
+                }
+            }
+        }
+
+        // Default: keep original call
+        caps.get(0).unwrap().as_str().to_string()
+    })
+    .into_owned()
+}
+
+use common::proto::komp_ac::table_definition::ColumnDefinition;
+
+// Converts row data boolean values to Steel script format during context initialization.
+pub async fn convert_row_data_for_steel(
+    db_pool: &PgPool,
+    schema_id: i64,
+    table_name: &str,
+    row_data: &mut HashMap<String, String>,
+) -> Result<(), sqlx::Error> {
+    let table_def = sqlx::query!(
+        r#"
+        SELECT columns FROM table_definitions
+        WHERE schema_id = $1 AND table_name = $2
+        "#,
+        schema_id,
+        table_name
+    )
+    .fetch_optional(db_pool)
+    .await?
+    .ok_or_else(|| sqlx::Error::RowNotFound)?;
+
+    // Parse column definitions to identify boolean columns
+    if let Ok(columns) = serde_json::from_value::<Vec<ColumnDefinition>>(table_def.columns) {
+        for col_def in columns {
+            let normalized_type =
+                col_def.field_type.to_uppercase().split('(').next().unwrap().to_string();
+
+            if normalized_type == "BOOLEAN" || normalized_type == "BOOL" {
+                if let Some(value) = row_data.get_mut(&col_def.name) {
+                    *value = match value.to_lowercase().as_str() {
+                        "true" | "t" | "1" | "yes" | "on" => "#true".to_string(),
+                        "false" | "f" | "0" | "no" | "off" => "#false".to_string(),
+                        _ => value.clone(),
+                    };
+                }
+            }
+        }
+    }
+
+    Ok(())
+}
+
 pub async fn create_steel_context_with_boolean_conversion(
     current_table: String,
     schema_id: i64,
@@ -36,7 +111,6 @@ pub async fn create_steel_context_with_boolean_conversion(
     mut row_data: HashMap<String, String>,
     db_pool: Arc<PgPool>,
 ) -> Result<SteelContext, ExecutionError> {
-    // Convert boolean values in row_data to Steel format
     convert_row_data_for_steel(&db_pool, schema_id, &current_table, &mut row_data)
         .await
         .map_err(|e| {
@@ -53,7 +127,6 @@ pub async fn create_steel_context_with_boolean_conversion(
     })
 }
 
-/// Executes a Steel script with database context and type-safe result processing.
 pub async fn execute_script(
     script: String,
     target_type: &str,
@@ -65,42 +138,40 @@ pub async fn execute_script(
 ) -> Result<Value, ExecutionError> {
     let mut vm = Engine::new();
 
-    // Create execution context with proper boolean value conversion
+    // Upgrade to with_index based on FK presence in the posted data
+    let script = auto_promote_with_index(&script, &current_table, &row_data);
+
     let context = create_steel_context_with_boolean_conversion(
-        current_table,
+        current_table.clone(),
         schema_id,
         schema_name,
-        row_data,
+        row_data.clone(),
         db_pool.clone(),
-    ).await?;
+    )
+    .await?;
 
     let context = Arc::new(context);
 
-    // Register database access functions
     register_steel_functions(&mut vm, context.clone());
-
-    // Register decimal math operations
     register_decimal_math_functions(&mut vm);
 
-    // Register row data as variables in the Steel VM for get-var access
     let mut define_script = String::new();
-
     for (key, value) in &context.row_data {
-        // Register only bare variable names for get-var access
         define_script.push_str(&format!("(define {} \"{}\")\n", key, value));
     }
 
-    // Execute variable definitions if any exist
     if !define_script.is_empty() {
         vm.compile_and_run_raw_program(define_script)
-            .map_err(|e| ExecutionError::RuntimeError(format!("Failed to register variables: {}", e)))?;
+            .map_err(|e| ExecutionError::RuntimeError(format!(
+                "Failed to register variables: {}",
+                e
+            )))?;
     }
 
-    // Also register variables using the decimal registry as backup method
     FunctionRegistry::register_variables(&mut vm, context.row_data.clone());
 
-    // Execute the main script
-    let results = vm.compile_and_run_raw_program(script.clone())
+    let results = vm
+        .compile_and_run_raw_program(script.clone())
         .map_err(|e| {
             error!("Steel script execution failed: {}", e);
             error!("Script was: {}", script);
@@ -108,22 +179,22 @@ pub async fn execute_script(
             ExecutionError::RuntimeError(e.to_string())
         })?;
 
-    // Convert results to the requested target type
     match target_type {
         "STRINGS" => process_string_results(results),
-        _ => Err(ExecutionError::UnsupportedType(target_type.into()))
+        _ => Err(ExecutionError::UnsupportedType(target_type.into())),
     }
 }
 
-/// Registers Steel functions for database access within the VM context.
 fn register_steel_functions(vm: &mut Engine, context: Arc<SteelContext>) {
     debug!("Registering Steel functions with context");
 
-    // Register column access function for current and related tables
     vm.register_fn("steel_get_column", {
         let ctx = context.clone();
         move |table: String, column: String| {
-            debug!("steel_get_column called with table: '{}', column: '{}'", table, column);
+            debug!(
+                "steel_get_column called with table: '{}', column: '{}'",
+                table, column
+            );
             ctx.steel_get_column(&table, &column)
                 .map_err(|e| {
                     error!("steel_get_column failed: {:?}", e);
@@ -132,11 +203,13 @@ fn register_steel_functions(vm: &mut Engine, context: Arc<SteelContext>) {
         }
     });
 
-    // Register indexed column access for comma-separated values
     vm.register_fn("steel_get_column_with_index", {
         let ctx = context.clone();
         move |table: String, index: i64, column: String| {
-            debug!("steel_get_column_with_index called with table: '{}', index: {}, column: '{}'", table, index, column);
+            debug!(
+                "steel_get_column_with_index called with table: '{}', index: {}, column: '{}'",
+                table, index, column
+            );
             ctx.steel_get_column_with_index(&table, index, &column)
                 .map_err(|e| {
                     error!("steel_get_column_with_index failed: {:?}", e);
@@ -145,27 +218,23 @@ fn register_steel_functions(vm: &mut Engine, context: Arc<SteelContext>) {
         }
     });
 
-    // Register safe SQL query execution
     vm.register_fn("steel_query_sql", {
         let ctx = context.clone();
         move |query: String| {
             debug!("steel_query_sql called with query: '{}'", query);
-            ctx.steel_query_sql(&query)
-                .map_err(|e| {
-                    error!("steel_query_sql failed: {:?}", e);
-                    e.to_string()
-                })
+            ctx.steel_query_sql(&query).map_err(|e| {
+                error!("steel_query_sql failed: {:?}", e);
+                e.to_string()
+            })
         }
     });
 }
 
-/// Registers decimal mathematics functions in the Steel VM.
 fn register_decimal_math_functions(vm: &mut Engine) {
     debug!("Registering decimal math functions");
     FunctionRegistry::register_all(vm);
 }
 
-/// Processes Steel script results into string format for consistent output.
 fn process_string_results(results: Vec<SteelVal>) -> Result<Value, ExecutionError> {
     let mut strings = Vec::new();
 
@@ -178,7 +247,7 @@ fn process_string_results(results: Vec<SteelVal>) -> Result<Value, ExecutionErro
             _ => {
                 error!("Unexpected result type: {:?}", result);
                 return Err(ExecutionError::TypeConversionError(
-                    format!("Expected string-convertible type, got {:?}", result)
+                    format!("Expected string-convertible type, got {:?}", result),
                 ));
             }
         };

server/src/steel/server/functions.rs

@@ -1,5 +1,6 @@
 // src/steel/server/functions.rs
 
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use steel::rvals::SteelVal;
 use sqlx::PgPool;
 use std::collections::HashMap;
@@ -21,10 +22,8 @@ pub enum FunctionError {
     ProhibitedTypeAccess(String),
 }
 
-/// Data types that Steel scripts are prohibited from accessing for security reasons
 const PROHIBITED_TYPES: &[&str] = &["BIGINT", "DATE", "TIMESTAMPTZ"];
 
-/// Execution context for Steel scripts with database access capabilities.
 #[derive(Clone)]
 pub struct SteelContext {
     pub current_table: String,
@@ -35,26 +34,11 @@ pub struct SteelContext {
 }
 
 impl SteelContext {
-    /// Resolves a base table name to its full qualified name in the current schema.
-    /// Used for foreign key relationship traversal in Steel scripts.
-    pub async fn get_related_table_name(&self, base_name: &str) -> Result<String, FunctionError> {
-        let table_def = sqlx::query!(
-            r#"SELECT table_name FROM table_definitions
-               WHERE schema_id = $1 AND table_name LIKE $2"#,
-            self.schema_id,
-            format!("%_{}", base_name)
-        )
-        .fetch_optional(&*self.db_pool)
-        .await
-        .map_err(|e| FunctionError::DatabaseError(e.to_string()))?
-        .ok_or_else(|| FunctionError::TableNotFound(base_name.to_string()))?;
-
-        Ok(table_def.table_name)
-    }
-
-    /// Retrieves the SQL data type for a specific column in a table.
-    /// Parses the JSON column definitions to find type information.
-    async fn get_column_type(&self, table_name: &str, column_name: &str) -> Result<String, FunctionError> {
+    async fn get_column_type(
+        &self,
+        table_name: &str,
+        column_name: &str,
+    ) -> Result<String, FunctionError> {
         let table_def = sqlx::query!(
             r#"SELECT columns FROM table_definitions
                WHERE schema_id = $1 AND table_name = $2"#,
@@ -66,49 +50,43 @@ impl SteelContext {
         .map_err(|e| FunctionError::DatabaseError(e.to_string()))?
         .ok_or_else(|| FunctionError::TableNotFound(table_name.to_string()))?;
 
-        let columns: Vec<String> = serde_json::from_value(table_def.columns)
-            .map_err(|e| FunctionError::DatabaseError(format!("Invalid column data: {}", e)))?;
+        let columns: Vec<ColumnDefinition> = serde_json::from_value(table_def.columns)
+            .map_err(|e| FunctionError::DatabaseError(format!(
+                "Invalid column data: {}",
+                e
+            )))?;
 
-        // Parse column definitions to find the requested column type
-        for column_def in columns {
-            let mut parts = column_def.split_whitespace();
-            if let (Some(name), Some(data_type)) = (parts.next(), parts.next()) {
-                let column_name_clean = name.trim_matches('"');
-                if column_name_clean == column_name {
-                    return Ok(data_type.to_string());
-                }
+        for col_def in columns {
+            if col_def.name == column_name {
+                return Ok(col_def.field_type.to_uppercase());
             }
         }
 
         Err(FunctionError::ColumnNotFound(format!(
             "Column '{}' not found in table '{}'",
-            column_name,
-            table_name
+            column_name, table_name
         )))
     }
 
-    /// Converts database values to Steel script format based on column type.
-    /// Currently handles boolean conversion to Steel's #true/#false syntax.
     fn convert_value_to_steel_format(&self, value: &str, column_type: &str) -> String {
         let normalized_type = normalize_data_type(column_type);
 
         match normalized_type.as_str() {
-            "BOOLEAN" | "BOOL" => {
-                // Convert database boolean representations to Steel boolean syntax
-                match value.to_lowercase().as_str() {
-                    "true" | "t" | "1" | "yes" | "on" => "#true".to_string(),
-                    "false" | "f" | "0" | "no" | "off" => "#false".to_string(),
-                    _ => value.to_string(), // Return as-is if not a recognized boolean
-                }
-            }
+            "BOOLEAN" | "BOOL" => match value.to_lowercase().as_str() {
+                "true" | "t" | "1" | "yes" | "on" => "#true".to_string(),
+                "false" | "f" | "0" | "no" | "off" => "#false".to_string(),
+                _ => value.to_string(),
+            },
             "INTEGER" => value.to_string(),
-            _ => value.to_string(), // Return as-is for other types
+            _ => value.to_string(),
         }
     }
 
-    /// Validates that a column type is allowed for Steel script access.
-    /// Returns the column type if validation passes, error if prohibited.
-    async fn validate_column_type_and_get_type(&self, table_name: &str, column_name: &str) -> Result<String, FunctionError> {
+    async fn validate_column_type_and_get_type(
+        &self,
+        table_name: &str,
+        column_name: &str,
+    ) -> Result<String, FunctionError> {
         let column_type = self.get_column_type(table_name, column_name).await?;
 
         if is_prohibited_type(&column_type) {
@@ -124,15 +102,13 @@ impl SteelContext {
         Ok(column_type)
     }
 
-    /// Retrieves column value from current table or related tables via foreign keys.
-    /// 
-    /// # Behavior
-    /// - Current table: Returns value directly from row_data with type conversion
-    /// - Related table: Follows foreign key relationship and queries database
-    /// - All accesses are subject to prohibited type validation
-    pub fn steel_get_column(&self, table: &str, column: &str) -> Result<SteelVal, SteelVal> {
+    pub fn steel_get_column(
+        &self,
+        table: &str,
+        column: &str,
+    ) -> Result<SteelVal, SteelVal> {
         if table == self.current_table {
-            // Access current table data with type validation
+            // current table
             let column_type = tokio::task::block_in_place(|| {
                 let handle = tokio::runtime::Handle::current();
                 handle.block_on(async {
@@ -145,70 +121,112 @@ impl SteelContext {
                 Err(e) => return Err(SteelVal::StringV(e.to_string().into())),
             };
 
-            return self.row_data.get(column)
+            return self
+                .row_data
+                .get(column)
                 .map(|v| {
-                    let converted_value = self.convert_value_to_steel_format(v, &column_type);
-                    SteelVal::StringV(converted_value.into())
+                    let converted =
+                        self.convert_value_to_steel_format(v, &column_type);
+                    SteelVal::StringV(converted.into())
                 })
-                .ok_or_else(|| SteelVal::StringV(format!("Column {} not found", column).into()));
+                .ok_or_else(|| {
+                    SteelVal::StringV(
+                        format!("Column {} not found", column).into(),
+                    )
+                });
         }
 
-        // Access related table via foreign key relationship
-        let base_name = table.split_once('_')
-            .map(|(_, rest)| rest)
-            .unwrap_or(table);
-
-        let fk_column = format!("{}_id", base_name);
-        let fk_value = self.row_data.get(&fk_column)
-            .ok_or_else(|| SteelVal::StringV(format!("Foreign key {} not found", fk_column).into()))?;
-
+        // Cross-table via FK: use exact table name FK convention: "<table>_id"
         let result = tokio::task::block_in_place(|| {
             let handle = tokio::runtime::Handle::current();
             handle.block_on(async {
-                let actual_table = self.get_related_table_name(base_name).await
-                    .map_err(|e| SteelVal::StringV(e.to_string().into()))?;
+                let fk_column = format!("{}_id", table);
+                let fk_value = self
+                    .row_data
+                    .get(&fk_column)
+                    .ok_or_else(|| {
+                        FunctionError::ForeignKeyNotFound(format!(
+                            "Foreign key column '{}' not found on '{}'",
+                            fk_column, self.current_table
+                        ))
+                    })?;
 
-                // Validate column type and get type information
-                let column_type = self.validate_column_type_and_get_type(&actual_table, column).await
-                    .map_err(|e| SteelVal::StringV(e.to_string().into()))?;
+                let column_type =
+                    self.validate_column_type_and_get_type(table, column)
+                        .await?;
 
-                // Query the related table for the column value
-                let raw_value = sqlx::query_scalar::<_, String>(
-                    &format!("SELECT {} FROM \"{}\".\"{}\" WHERE id = $1", column, self.schema_name, actual_table)
+                let raw_value = sqlx::query_scalar::<_, String>(&format!(
+                    "SELECT \"{}\" FROM \"{}\".\"{}\" WHERE id = $1",
+                    column, self.schema_name, table
+                ))
+                .bind(
+                    fk_value
+                        .parse::<i64>()
+                        .map_err(|_| {
+                            FunctionError::DatabaseError(
+                                "Invalid foreign key format".into(),
+                            )
+                        })?,
                 )
-                .bind(fk_value.parse::<i64>().map_err(|_|
-                    SteelVal::StringV("Invalid foreign key format".into()))?)
                 .fetch_one(&*self.db_pool)
                 .await
-                .map_err(|e| SteelVal::StringV(e.to_string().into()))?;
+                .map_err(|e| FunctionError::DatabaseError(e.to_string()))?;
 
-                // Convert to appropriate Steel format
-                let converted_value = self.convert_value_to_steel_format(&raw_value, &column_type);
-                Ok(converted_value)
+                let converted =
+                    self.convert_value_to_steel_format(&raw_value, &column_type);
+                Ok::<String, FunctionError>(converted)
             })
         });
 
-        result.map(|v| SteelVal::StringV(v.into()))
+        match result {
+            Ok(v) => Ok(SteelVal::StringV(v.into())),
+            Err(e) => Err(SteelVal::StringV(e.to_string().into())),
+        }
     }
 
-    /// Retrieves a specific indexed element from a comma-separated column value.
-    /// Useful for accessing elements from array-like string representations.
     pub fn steel_get_column_with_index(
         &self,
         table: &str,
         index: i64,
-        column: &str
+        column: &str,
     ) -> Result<SteelVal, SteelVal> {
-        // Get the full value with proper type conversion
-        let value = self.steel_get_column(table, column)?;
+        // Cross-table: interpret 'index' as the row id to fetch directly
+        if table != self.current_table {
+            let result = tokio::task::block_in_place(|| {
+                let handle = tokio::runtime::Handle::current();
+                handle.block_on(async {
+                    let column_type =
+                        self.validate_column_type_and_get_type(table, column)
+                            .await?;
 
+                    let raw_value = sqlx::query_scalar::<_, String>(&format!(
+                        "SELECT \"{}\" FROM \"{}\".\"{}\" WHERE id = $1",
+                        column, self.schema_name, table
+                    ))
+                    .bind(index)
+                    .fetch_one(&*self.db_pool)
+                    .await
+                    .map_err(|e| FunctionError::DatabaseError(e.to_string()))?;
+
+                    let converted = self
+                        .convert_value_to_steel_format(&raw_value, &column_type);
+                    Ok::<String, FunctionError>(converted)
+                })
+            });
+
+            return match result {
+                Ok(v) => Ok(SteelVal::StringV(v.into())),
+                Err(e) => Err(SteelVal::StringV(e.to_string().into())),
+            };
+        }
+
+        // Current table: existing behavior (index in comma-separated string)
+        let value = self.steel_get_column(table, column)?;
         if let SteelVal::StringV(s) = value {
             let parts: Vec<_> = s.split(',').collect();
-
             if let Some(part) = parts.get(index as usize) {
-                let trimmed_part = part.trim();
+                let trimmed = part.trim();
 
-                // Apply type conversion to the indexed part based on original column type
                 let column_type = tokio::task::block_in_place(|| {
                     let handle = tokio::runtime::Handle::current();
                     handle.block_on(async {
@@ -218,40 +236,35 @@ impl SteelContext {
 
                 match column_type {
                     Ok(ct) => {
-                        let converted_part = self.convert_value_to_steel_format(trimmed_part, &ct);
-                        Ok(SteelVal::StringV(converted_part.into()))
-                    }
-                    Err(_) => {
-                        // If type cannot be determined, return value as-is
-                        Ok(SteelVal::StringV(trimmed_part.into()))
+                        let converted =
+                            self.convert_value_to_steel_format(trimmed, &ct);
+                        Ok(SteelVal::StringV(converted.into()))
                     }
+                    Err(_) => Ok(SteelVal::StringV(trimmed.into())),
                 }
             } else {
                 Err(SteelVal::StringV("Index out of bounds".into()))
             }
         } else {
-            Err(SteelVal::StringV("Expected comma-separated string".into()))
+            Err(SteelVal::StringV(
+                "Expected comma-separated string".into(),
+            ))
         }
     }
 
-    /// Executes read-only SQL queries from Steel scripts with safety restrictions.
-    /// 
-    /// # Security Features
-    /// - Only SELECT, SHOW, and EXPLAIN queries allowed
-    /// - Prohibited column type access validation
-    /// - Returns first column of all rows as comma-separated string
     pub fn steel_query_sql(&self, query: &str) -> Result<SteelVal, SteelVal> {
         if !is_read_only_query(query) {
-            return Err(SteelVal::StringV(
-                "Only SELECT queries are allowed".into()
-            ));
+            return Err(SteelVal::StringV("Only SELECT queries are allowed".into()));
         }
 
         if contains_prohibited_column_access(query) {
-            return Err(SteelVal::StringV(format!(
-                "SQL query may access prohibited column types. Steel scripts cannot access columns of type: {}",
-                PROHIBITED_TYPES.join(", ")
-            ).into()));
+            return Err(SteelVal::StringV(
+                format!(
+                    "SQL query may access prohibited column types. Steel scripts cannot access columns of type: {}",
+                    PROHIBITED_TYPES.join(", ")
+                )
+                .into(),
+            ));
         }
 
         let pool = self.db_pool.clone();
@@ -266,7 +279,8 @@ impl SteelContext {
 
                 let mut results = Vec::new();
                 for row in rows {
-                    let val: String = row.try_get(0)
+                    let val: String = row
+                        .try_get(0)
                         .map_err(|e| SteelVal::StringV(e.to_string().into()))?;
                     results.push(val);
                 }
@@ -279,85 +293,30 @@ impl SteelContext {
     }
 }
 
-/// Checks if a data type is prohibited for Steel script access.
 fn is_prohibited_type(data_type: &str) -> bool {
     let normalized_type = normalize_data_type(data_type);
-    PROHIBITED_TYPES.iter().any(|&prohibited| normalized_type.starts_with(prohibited))
+    PROHIBITED_TYPES
+        .iter()
+        .any(|&prohibited| normalized_type.starts_with(prohibited))
 }
 
-/// Normalizes data type strings for consistent comparison.
-/// Handles variations like NUMERIC(10,2) by extracting base type.
 fn normalize_data_type(data_type: &str) -> String {
-    data_type.to_uppercase()
-        .split('(')  // Remove precision/scale from NUMERIC(x,y)
+    data_type
+        .to_uppercase()
+        .split('(')
         .next()
         .unwrap_or(data_type)
         .trim()
         .to_string()
 }
 
-/// Performs basic heuristic check for prohibited column type access in SQL queries.
-/// Looks for common patterns that might indicate access to restricted types.
 fn contains_prohibited_column_access(query: &str) -> bool {
     let query_upper = query.to_uppercase();
-
-    let patterns = [
-        "EXTRACT(", // Common with DATE/TIMESTAMPTZ
-        "DATE_PART(", // Common with DATE/TIMESTAMPTZ
-        "::DATE",
-        "::TIMESTAMPTZ",
-        "::BIGINT",
-    ];
-
-    patterns.iter().any(|pattern| query_upper.contains(pattern))
+    let patterns = ["EXTRACT(", "DATE_PART(", "::DATE", "::TIMESTAMPTZ", "::BIGINT"];
+    patterns.iter().any(|p| query_upper.contains(p))
 }
 
-/// Validates that a query is read-only and safe for Steel script execution.
 fn is_read_only_query(query: &str) -> bool {
     let query = query.trim_start().to_uppercase();
-    query.starts_with("SELECT") ||
-    query.starts_with("SHOW") ||
-    query.starts_with("EXPLAIN")
-}
-
-/// Converts row data boolean values to Steel script format during context initialization.
-pub async fn convert_row_data_for_steel(
-    db_pool: &PgPool,
-    schema_id: i64,
-    table_name: &str,
-    row_data: &mut HashMap<String, String>,
-) -> Result<(), sqlx::Error> {
-    let table_def = sqlx::query!(
-        r#"SELECT columns FROM table_definitions
-           WHERE schema_id = $1 AND table_name = $2"#,
-        schema_id,
-        table_name
-    )
-    .fetch_optional(db_pool)
-    .await?
-    .ok_or_else(|| sqlx::Error::RowNotFound)?;
-
-    // Parse column definitions to identify boolean columns for conversion
-    if let Ok(columns) = serde_json::from_value::<Vec<String>>(table_def.columns) {
-        for column_def in columns {
-            let mut parts = column_def.split_whitespace();
-            if let (Some(name), Some(data_type)) = (parts.next(), parts.next()) {
-                let column_name = name.trim_matches('"');
-                let normalized_type = normalize_data_type(data_type);
-
-                if normalized_type == "BOOLEAN" || normalized_type == "BOOL" {
-                    if let Some(value) = row_data.get_mut(column_name) {
-                        // Convert boolean value to Steel format
-                        *value = match value.to_lowercase().as_str() {
-                            "true" | "t" | "1" | "yes" | "on" => "#true".to_string(),
-                            "false" | "f" | "0" | "no" | "off" => "#false".to_string(),
-                            _ => value.clone(), // Keep original if not recognized
-                        };
-                    }
-                }
-            }
-        }
-    }
-
-    Ok(())
+    query.starts_with("SELECT") || query.starts_with("SHOW") || query.starts_with("EXPLAIN")
 }

server/src/table_definition/handlers/post_table_definition.rs

@@ -4,22 +4,8 @@ use tonic::Status;
 use sqlx::{PgPool, Transaction, Postgres};
 use serde_json::json;
 use common::proto::komp_ac::table_definition::{PostTableDefinitionRequest, TableDefinitionResponse};
-
-// TODO CRITICAL add decimal with optional precision"
-const PREDEFINED_FIELD_TYPES: &[(&str, &str)] = &[
-    ("text", "TEXT"),
-    ("string", "TEXT"),
-    ("boolean", "BOOLEAN"),
-    ("timestamp", "TIMESTAMPTZ"),
-    ("timestamptz", "TIMESTAMPTZ"),
-    ("time", "TIMESTAMPTZ"),
-    ("money", "NUMERIC(14, 4)"),
-    ("integer", "INTEGER"),
-    ("int", "INTEGER"),
-    ("biginteger", "BIGINT"),
-    ("bigint", "BIGINT"),
-    ("date", "DATE"),
-];
+use common::proto::komp_ac::table_definition::ColumnDefinition;
+use crate::table_definition::models::map_field_type;
 
 // NEW: Helper function to provide detailed error messages
 fn validate_identifier_format(s: &str, identifier_type: &str) -> Result<(), Status> {
@@ -58,116 +44,6 @@ fn validate_identifier_format(s: &str, identifier_type: &str) -> Result<(), Stat
     Ok(())
 }
 
-fn validate_decimal_number_format(num_str: &str, param_name: &str) -> Result<(), Status> {
-    if num_str.is_empty() {
-        return Err(Status::invalid_argument(format!(
-            "{} cannot be empty",
-            param_name
-        )));
-    }
-
-    // Check for explicit signs
-    if num_str.starts_with('+') || num_str.starts_with('-') {
-        return Err(Status::invalid_argument(format!(
-            "{} cannot have explicit positive or negative signs",
-            param_name
-        )));
-    }
-
-    // Check for decimal points
-    if num_str.contains('.') {
-        return Err(Status::invalid_argument(format!(
-            "{} must be a whole number (no decimal points)",
-            param_name
-        )));
-    }
-
-    // Check for leading zeros (but allow "0" itself)
-    if num_str.len() > 1 && num_str.starts_with('0') {
-        let trimmed = num_str.trim_start_matches('0');
-        let suggestion = if trimmed.is_empty() { "0" } else { trimmed };
-        return Err(Status::invalid_argument(format!(
-            "{} cannot have leading zeros (use '{}' instead of '{}')",
-            param_name,
-            suggestion,
-            num_str
-        )));
-    }
-
-    // Check that all characters are digits
-    if !num_str.chars().all(|c| c.is_ascii_digit()) {
-        return Err(Status::invalid_argument(format!(
-            "{} contains invalid characters. Only digits 0-9 are allowed",
-            param_name
-        )));
-    }
-
-    Ok(())
-}
-
-fn map_field_type(field_type: &str) -> Result<String, Status> {
-    let lower_field_type = field_type.to_lowercase();
-
-    // Special handling for "decimal(precision, scale)"
-    if lower_field_type.starts_with("decimal(") && lower_field_type.ends_with(')') {
-        // Extract the part inside the parentheses, e.g., "10, 2"
-        let args = lower_field_type
-            .strip_prefix("decimal(")
-            .and_then(|s| s.strip_suffix(')'))
-            .unwrap_or(""); // Should always succeed due to the checks above
-
-        // Split into precision and scale parts
-        if let Some((p_str, s_str)) = args.split_once(',') {
-            let precision_str = p_str.trim();
-            let scale_str = s_str.trim();
-
-            // NEW: Validate format BEFORE parsing
-            validate_decimal_number_format(precision_str, "precision")?;
-            validate_decimal_number_format(scale_str, "scale")?;
-
-            // Parse precision, returning an error if it's not a valid number
-            let precision = precision_str.parse::<u32>().map_err(|_| {
-                Status::invalid_argument("Invalid precision in decimal type")
-            })?;
-
-            // Parse scale, returning an error if it's not a valid number
-            let scale = scale_str.parse::<u32>().map_err(|_| {
-                Status::invalid_argument("Invalid scale in decimal type")
-            })?;
-
-            // Add validation based on PostgreSQL rules
-            if precision < 1 {
-                return Err(Status::invalid_argument("Precision must be at least 1"));
-            }
-            if scale > precision {
-                return Err(Status::invalid_argument(
-                    "Scale cannot be greater than precision",
-                ));
-            }
-
-            // If everything is valid, build and return the NUMERIC type string
-            return Ok(format!("NUMERIC({}, {})", precision, scale));
-        } else {
-            // The format was wrong, e.g., "decimal(10)" or "decimal()"
-            return Err(Status::invalid_argument(
-                "Invalid decimal format. Expected: decimal(precision, scale)",
-            ));
-        }
-    }
-
-    // If not a decimal, fall back to the predefined list
-    PREDEFINED_FIELD_TYPES
-        .iter()
-        .find(|(key, _)| *key == lower_field_type.as_str())
-        .map(|(_, sql_type)| sql_type.to_string()) // Convert to an owned String
-        .ok_or_else(|| {
-            Status::invalid_argument(format!(
-                "Invalid field type: {}",
-                field_type
-            ))
-        })
-}
-
 fn is_invalid_table_name(table_name: &str) -> bool {
     table_name.ends_with("_id") ||
     table_name == "id" ||
@@ -299,34 +175,48 @@ async fn execute_table_definition(
         links.push((linked_id, link.required));
     }
 
-    let mut columns = Vec::new();
+    let mut stored_columns = Vec::new();
+    let mut sql_columns = Vec::new();
+
     for col_def in request.columns.drain(..) {
         let col_name = col_def.name.trim().to_string();
         validate_identifier_format(&col_name, "Column name")?;
 
         if col_name.ends_with("_id") || col_name == "id" || col_name == "deleted" || col_name == "created_at" {
             return Err(Status::invalid_argument(format!(
-                "Column name '{}' cannot be 'id', 'deleted', 'created_at' or end with '_id'",
-                col_name
+                        "Column name '{}' cannot be 'id', 'deleted', 'created_at' or end with '_id'",
+                        col_name
             )));
         }
 
         let sql_type = map_field_type(&col_def.field_type)?;
-        columns.push(format!("\"{}\" {}", col_name, sql_type));
+        sql_columns.push(format!("\"{}\" {}", col_name, sql_type));
+
+        // push the proto type (serde serializable)
+        stored_columns.push(ColumnDefinition {
+            name: col_name,
+            field_type: col_def.field_type,
+        });
     }
 
+    // Indexes
+    let mut stored_indexes = Vec::new();
     let mut indexes = Vec::new();
     for idx in request.indexes.drain(..) {
         let idx_name = idx.trim().to_string();
         validate_identifier_format(&idx_name, "Index name")?;
 
-        if !columns.iter().any(|c| c.starts_with(&format!("\"{}\"", idx_name))) {
-            return Err(Status::invalid_argument(format!("Index column '{}' not found", idx_name)));
+        if !sql_columns.iter().any(|c| c.starts_with(&format!("\"{}\"", idx_name))) {
+            return Err(Status::invalid_argument(format!(
+                        "Index column '{}' not found", idx_name
+            )));
         }
+
+        stored_indexes.push(idx_name.clone());
         indexes.push(idx_name);
     }
 
-    let (create_sql, index_sql) = generate_table_sql(tx, &profile_name, &table_name, &columns, &indexes, &links).await?;
+    let (create_sql, index_sql) = generate_table_sql(tx, &profile_name, &table_name, &sql_columns, &indexes, &links).await?;
 
     // Use schema_id instead of profile_id
     let table_def = sqlx::query!(
@@ -336,8 +226,8 @@ async fn execute_table_definition(
            RETURNING id"#,
         schema.id,
         &table_name,
-        json!(columns),
-        json!(indexes)
+        serde_json::to_value(&stored_columns).unwrap(),
+        serde_json::to_value(&stored_indexes).unwrap()
     )
     .fetch_one(&mut **tx)
     .await
@@ -351,7 +241,7 @@ async fn execute_table_definition(
         Status::internal(format!("Database error: {}", e))
     })?;
 
-    for col_def in &columns {
+    for col_def in &sql_columns {
         // Column string looks like "\"name\" TYPE", split out identifier
         let col_name = col_def.split_whitespace().next().unwrap_or("");
         let clean_col = col_name.trim_matches('"');

server/src/table_definition/mod.rs

@@ -2,3 +2,6 @@
 
 pub mod models;
 pub mod handlers;
+pub mod repo;
+
+pub use repo::*;

server/src/table_definition/models.rs

@@ -0,0 +1,91 @@
+// src/table_definition/models.rs
+use tonic::Status;
+
+/// Predefined static field mappings
+// TODO CRITICAL add decimal with optional precision"
+pub const PREDEFINED_FIELD_TYPES: &[(&str, &str)] = &[
+    ("text", "TEXT"),
+    ("string", "TEXT"),
+    ("boolean", "BOOLEAN"),
+    ("timestamp", "TIMESTAMPTZ"),
+    ("timestamptz", "TIMESTAMPTZ"),
+    ("time", "TIMESTAMPTZ"),
+    ("money", "NUMERIC(14, 4)"),
+    ("integer", "INTEGER"),
+    ("int", "INTEGER"),
+    ("biginteger", "BIGINT"),
+    ("bigint", "BIGINT"),
+    ("date", "DATE"),
+];
+
+/// reusable decimal number validation
+pub fn validate_decimal_number_format(num_str: &str, param_name: &str) -> Result<(), Status> {
+    if num_str.is_empty() {
+        return Err(Status::invalid_argument(format!("{} cannot be empty", param_name)));
+    }
+    if num_str.starts_with('+') || num_str.starts_with('-') {
+        return Err(Status::invalid_argument(format!(
+            "{} cannot have explicit positive/negative signs", param_name
+        )));
+    }
+    if num_str.contains('.') {
+        return Err(Status::invalid_argument(format!(
+            "{} must be a whole number (no decimal point)", param_name
+        )));
+    }
+    if num_str.len() > 1 && num_str.starts_with('0') {
+        let trimmed = num_str.trim_start_matches('0');
+        let suggestion = if trimmed.is_empty() { "0" } else { trimmed };
+        return Err(Status::invalid_argument(format!(
+            "{} cannot have leading zeros (use '{}' instead of '{}')",
+            param_name, suggestion, num_str
+        )));
+    }
+    if !num_str.chars().all(|c| c.is_ascii_digit()) {
+        return Err(Status::invalid_argument(format!(
+            "{} contains invalid characters. Only digits allowed", param_name
+        )));
+    }
+    Ok(())
+}
+
+/// reusable field type mapper
+pub fn map_field_type(field_type: &str) -> Result<String, Status> {
+    let lower_field_type = field_type.to_lowercase();
+
+    if lower_field_type.starts_with("decimal(") && lower_field_type.ends_with(')') {
+        let args = lower_field_type.strip_prefix("decimal(").unwrap()
+            .strip_suffix(')').unwrap();
+
+        if let Some((p_str, s_str)) = args.split_once(',') {
+            let precision_str = p_str.trim();
+            let scale_str = s_str.trim();
+
+            validate_decimal_number_format(precision_str, "precision")?;
+            validate_decimal_number_format(scale_str, "scale")?;
+
+            let precision = precision_str.parse::<u32>()
+                .map_err(|_| Status::invalid_argument("Invalid precision"))?;
+            let scale = scale_str.parse::<u32>()
+                .map_err(|_| Status::invalid_argument("Invalid scale"))?;
+
+            if precision < 1 {
+                return Err(Status::invalid_argument("Precision must be >= 1"));
+            }
+            if scale > precision {
+                return Err(Status::invalid_argument("Scale cannot be > precision"));
+            }
+            return Ok(format!("NUMERIC({}, {})", precision, scale));
+        } else {
+            return Err(Status::invalid_argument(
+                "Invalid decimal format. Expected decimal(precision, scale)"
+            ));
+        }
+    }
+
+    PREDEFINED_FIELD_TYPES
+        .iter()
+        .find(|(key, _)| *key == lower_field_type.as_str())
+        .map(|(_, sql_type)| sql_type.to_string())
+        .ok_or_else(|| Status::invalid_argument(format!("Invalid field type: {}", field_type)))
+}

server/src/table_definition/repo.rs

@@ -0,0 +1,33 @@
+// src/table_definition/repo.rs
+use common::proto::komp_ac::table_definition::ColumnDefinition;
+use sqlx::PgPool;
+
+pub struct TableDefRow {
+    pub id: i64,
+    pub table_name: String,
+    pub columns: Vec<ColumnDefinition>,
+    pub indexes: Vec<String>,
+}
+
+pub async fn get_table_definition(
+    db: &PgPool,
+    id: i64,
+) -> Result<TableDefRow, anyhow::Error> {
+    let rec = sqlx::query!(
+        r#"
+        SELECT id, table_name, columns, indexes
+        FROM table_definitions
+        WHERE id = $1
+        "#,
+        id
+    )
+    .fetch_one(db)
+    .await?;
+
+    Ok(TableDefRow {
+        id: rec.id,
+        table_name: rec.table_name,
+        columns: serde_json::from_value(rec.columns)?, // 🔑
+        indexes: serde_json::from_value(rec.indexes)?,
+    })
+}

server/src/table_script/handlers/dependency_analyzer.rs

@@ -2,6 +2,7 @@
 
 use std::collections::HashMap;
 use tonic::Status;
+use serde::{Deserialize, Serialize};
 use serde_json::{json, Value};
 
 /// Represents the state of a node during dependency graph traversal.
@@ -40,18 +41,38 @@ impl DependencyType {
             DependencyType::SqlQuery { .. } => "sql_query",
         }
     }
+}
 
-    /// Generates context JSON for database storage.
-    pub fn context_json(&self) -> Value {
+/// Strongly-typed JSON for script_dependencies.context_info
+/// Using untagged so JSON stays minimal (no "type" field), and we can still
+/// deserialize it into a proper enum.
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(untagged)]
+pub enum ScriptDependencyContext {
+    ColumnAccess { column: String },
+    IndexedAccess { column: String, index: i64 },
+    SqlQuery { query_fragment: String },
+}
+
+impl DependencyType {
+    /// Convert this dependency into its JSON context struct.
+    pub fn to_context(&self) -> ScriptDependencyContext {
         match self {
             DependencyType::ColumnAccess { column } => {
-                json!({ "column": column })
+                ScriptDependencyContext::ColumnAccess {
+                    column: column.clone(),
+                }
             }
             DependencyType::IndexedAccess { column, index } => {
-                json!({ "column": column, "index": index })
+                ScriptDependencyContext::IndexedAccess {
+                    column: column.clone(),
+                    index: *index,
+                }
             }
             DependencyType::SqlQuery { query_fragment } => {
-                json!({ "query_fragment": query_fragment })
+                ScriptDependencyContext::SqlQuery {
+                    query_fragment: query_fragment.clone(),
+                }
             }
         }
     }
@@ -554,7 +575,7 @@ impl DependencyAnalyzer {
                 table_id,
                 target_id,
                 dep.dependency_type.as_str(),
-                dep.dependency_type.context_json()
+                serde_json::to_value(dep.dependency_type.to_context()).unwrap()
             )
             .execute(&mut **tx)
             .await

server/src/table_script/handlers/post_table_script.rs

@@ -4,6 +4,7 @@
 use tonic::Status;
 use sqlx::{PgPool, Error as SqlxError};
 use common::proto::komp_ac::table_script::{PostTableScriptRequest, TableScriptResponse};
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use serde_json::Value;
 use steel_decimal::SteelDecimal;
 use regex::Regex;
@@ -303,16 +304,12 @@ async fn validate_math_operations_column_types(
     let mut table_column_types: HashMap<String, HashMap<String, String>> = HashMap::new();
 
     for table_def in table_definitions {
-        let columns: Vec<String> = serde_json::from_value(table_def.columns)
+        let columns: Vec<ColumnDefinition> = serde_json::from_value(table_def.columns)
             .map_err(|e| Status::internal(format!("Invalid column data for table '{}': {}", table_def.table_name, e)))?;
 
         let mut column_types = HashMap::new();
-        for column_def in columns {
-            let mut parts = column_def.split_whitespace();
-            if let (Some(name), Some(data_type)) = (parts.next(), parts.next()) {
-                let column_name = name.trim_matches('"');
-                column_types.insert(column_name.to_string(), data_type.to_string());
-            }
+        for col_def in columns {
+            column_types.insert(col_def.name.clone(), col_def.field_type.clone());
         }
         table_column_types.insert(table_def.table_name, column_types);
     }
@@ -363,25 +360,13 @@ fn validate_target_column(
     }
 
     // Parse the columns JSON into a vector of strings
-    let columns: Vec<String> = serde_json::from_value(table_columns.clone())
+    let columns: Vec<ColumnDefinition> = serde_json::from_value(table_columns.clone())
         .map_err(|e| format!("Invalid column data: {}", e))?;
 
-    // Extract column names and types
-    let column_info: Vec<(&str, &str)> = columns
+    let column_type = columns
         .iter()
-        .filter_map(|c| {
-            let mut parts = c.split_whitespace();
-            let name = parts.next()?.trim_matches('"');
-            let data_type = parts.next()?;
-            Some((name, data_type))
-        })
-        .collect();
-
-    // Find the target column and return its type
-    let column_type = column_info
-        .iter()
-        .find(|(name, _)| *name == target)
-        .map(|(_, dt)| dt.to_string())
+        .find(|c| c.name == target)
+        .map(|c| c.field_type.clone())
         .ok_or_else(|| format!("Target column '{}' not defined in table '{}'", target, table_name))?;
 
     // Check if the target column type is prohibited
@@ -509,42 +494,29 @@ async fn validate_script_column_references(
 /// Validate that a referenced column doesn't have a prohibited type
 fn validate_referenced_column_type(table_name: &str, column_name: &str, table_columns: &Value) -> Result<(), String> {
     // Parse the columns JSON into a vector of strings
-    let columns: Vec<String> = serde_json::from_value(table_columns.clone())
-        .map_err(|e| format!("Invalid column data for table '{}': {}", table_name, e))?;
+let columns: Vec<ColumnDefinition> = serde_json::from_value(table_columns.clone())
+    .map_err(|e| format!("Invalid column data for table '{}': {}", table_name, e))?;
 
-    // Extract column names and types
-    let column_info: Vec<(&str, &str)> = columns
-        .iter()
-        .filter_map(|c| {
-            let mut parts = c.split_whitespace();
-            let name = parts.next()?.trim_matches('"');
-            let data_type = parts.next()?;
-            Some((name, data_type))
-        })
-        .collect();
-
-    // Find the referenced column and check its type
-    if let Some((_, column_type)) = column_info.iter().find(|(name, _)| *name == column_name) {
-        if is_prohibited_type(column_type) {
+    if let Some(col_def) = columns.iter().find(|c| c.name == column_name) {
+        if is_prohibited_type(&col_def.field_type) {
             return Err(format!(
-                "Script references column '{}' in table '{}' which has prohibited type '{}'. Steel scripts cannot access columns of type: {}",
-                column_name,
-                table_name,
-                column_type,
-                PROHIBITED_TYPES.join(", ")
+                    "Script references column '{}' in table '{}' which has prohibited type '{}'. Steel scripts cannot access columns of type: {}",
+                    column_name,
+                    table_name,
+                    col_def.field_type,
+                    PROHIBITED_TYPES.join(", ")
             ));
         }
 
-        // Log info for boolean columns
-        let normalized_type = normalize_data_type(column_type);
+        let normalized_type = normalize_data_type(&col_def.field_type);
         if normalized_type == "BOOLEAN" || normalized_type == "BOOL" {
             println!("Info: Script references boolean column '{}' in table '{}'. Values will be converted to Steel format (#true/#false)", column_name, table_name);
         }
     } else {
         return Err(format!(
-            "Script references column '{}' in table '{}' but this column does not exist",
-            column_name,
-            table_name
+                "Script references column '{}' in table '{}' but this column does not exist",
+                column_name,
+                table_name
         ));
     }
 

server/src/table_script/mod.rs

@@ -1,4 +1,7 @@
 // src/table_script/mod.rs
+
 pub mod handlers;
+pub mod repo;
 
 pub use handlers::*;
+pub use repo::*;

server/src/table_script/repo.rs

@@ -0,0 +1,49 @@
+// src/table_script/repo.rs
+use anyhow::Result;
+use sqlx::PgPool;
+
+use crate::table_script::handlers::dependency_analyzer::ScriptDependencyContext;
+
+#[derive(Debug, Clone)]
+pub struct ScriptDependencyRecord {
+    pub script_id: i64,
+    pub source_table_id: i64,
+    pub target_table_id: i64,
+    pub dependency_type: String,
+    pub context: Option<ScriptDependencyContext>,
+}
+
+pub async fn get_dependencies_for_script(
+    db: &PgPool,
+    script_id: i64,
+) -> Result<Vec<ScriptDependencyRecord>> {
+    let rows = sqlx::query!(
+        r#"
+        SELECT script_id, source_table_id, target_table_id, dependency_type, context_info
+        FROM script_dependencies
+        WHERE script_id = $1
+        ORDER BY source_table_id, target_table_id
+        "#,
+        script_id
+    )
+    .fetch_all(db)
+    .await?;
+
+    let mut out = Vec::new();
+    for r in rows {
+        let context = match r.context_info {
+            Some(value) => Some(serde_json::from_value::<ScriptDependencyContext>(value)?),
+            None => None,
+        };
+
+        out.push(ScriptDependencyRecord {
+            script_id: r.script_id,
+            source_table_id: r.source_table_id,
+            target_table_id: r.target_table_id,
+            dependency_type: r.dependency_type,
+            context,
+        });
+    }
+
+    Ok(out)
+}

server/src/tables_data/handlers/get_table_data.rs

@@ -1,9 +1,11 @@
 // src/tables_data/handlers/get_table_data.rs
 
+
 use tonic::Status;
 use sqlx::{PgPool, Row};
 use std::collections::HashMap;
 use common::proto::komp_ac::tables_data::{GetTableDataRequest, GetTableDataResponse};
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use crate::shared::schema_qualifier::qualify_table_name_for_data;
 
 pub async fn get_table_data(
@@ -39,17 +41,13 @@ pub async fn get_table_data(
     let table_def = table_def.ok_or_else(|| Status::not_found("Table not found"))?;
 
     // Parse user-defined columns from JSON
-    let columns_json: Vec<String> = serde_json::from_value(table_def.columns.clone())
+    let stored_columns: Vec<ColumnDefinition> = serde_json::from_value(table_def.columns.clone())
         .map_err(|e| Status::internal(format!("Column parsing error: {}", e)))?;
 
+    // Directly extract names, no split(" ") parsing anymore
     let mut user_columns = Vec::new();
-    for col_def in columns_json {
-        let parts: Vec<&str> = col_def.splitn(2, ' ').collect();
-        if parts.len() != 2 {
-            return Err(Status::internal("Invalid column format"));
-        }
-        let name = parts[0].trim_matches('"').to_string();
-        user_columns.push(name);
+    for col_def in stored_columns {
+        user_columns.push(col_def.name.trim().to_string());
     }
 
     // --- START OF FIX ---

server/src/tables_data/handlers/post_table_data.rs

@@ -5,6 +5,8 @@ use sqlx::{PgPool, Arguments};
 use sqlx::postgres::PgArguments;
 use chrono::{DateTime, Utc};
 use common::proto::komp_ac::tables_data::{PostTableDataRequest, PostTableDataResponse};
+use common::proto::komp_ac::table_definition::ColumnDefinition;
+use crate::table_definition::models::map_field_type;
 use std::collections::HashMap;
 use std::sync::Arc;
 use prost_types::value::Kind;
@@ -56,18 +58,16 @@ pub async fn post_table_data(
     let table_def = table_def.ok_or_else(|| Status::not_found("Table not found"))?;
 
     // Parse column definitions from JSON format
-    let columns_json: Vec<String> = serde_json::from_value(table_def.columns.clone())
+    let stored_columns: Vec<ColumnDefinition> = serde_json::from_value(table_def.columns.clone())
         .map_err(|e| Status::internal(format!("Column parsing error: {}", e)))?;
 
+    // convert ColumnDefinition -> (name, sql_type) using the same map_field_type logic
     let mut columns = Vec::new();
-    for col_def in columns_json {
-        let parts: Vec<&str> = col_def.splitn(2, ' ').collect();
-        if parts.len() != 2 {
-            return Err(Status::internal("Invalid column format"));
-        }
-        let name = parts[0].trim_matches('"').to_string();
-        let sql_type = parts[1].to_string();
-        columns.push((name, sql_type));
+    for col_def in stored_columns {
+        let col_name = col_def.name.trim().to_string();
+        let sql_type = map_field_type(&col_def.field_type)
+            .map_err(|e| Status::invalid_argument(format!("Invalid type for column '{}': {}", col_name, e)))?;
+        columns.push((col_name, sql_type));
     }
 
     // Build list of valid system columns (foreign keys and special columns)

server/src/tables_data/handlers/put_table_data.rs

@@ -5,6 +5,7 @@ use sqlx::{PgPool, Arguments, Row};
 use sqlx::postgres::PgArguments;
 use chrono::{DateTime, Utc};
 use common::proto::komp_ac::tables_data::{PutTableDataRequest, PutTableDataResponse};
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 
 use std::sync::Arc;
 use prost_types::value::Kind;
@@ -14,6 +15,7 @@ use std::collections::HashMap;
 
 use crate::steel::server::execution::{self, Value};
 use crate::indexer::{IndexCommand, IndexCommandData};
+use crate::table_definition::models::map_field_type;
 use tokio::sync::mpsc;
 use tracing::error;
 
@@ -56,19 +58,20 @@ pub async fn put_table_data(
     .map_err(|e| Status::internal(format!("Table lookup error: {}", e)))?
     .ok_or_else(|| Status::not_found("Table not found"))?;
 
-    // Parse column definitions from JSON format
-    let columns_json: Vec<String> = serde_json::from_value(table_def.columns.clone())
+    // Parse column definitions from JSON format (now ColumnDefinition objects)
+    let stored_columns: Vec<ColumnDefinition> = serde_json::from_value(table_def.columns.clone())
         .map_err(|e| Status::internal(format!("Column parsing error: {}", e)))?;
 
+    // Convert ColumnDefinition → (name, sql_type)
     let mut columns = Vec::new();
-    for col_def in columns_json {
-        let parts: Vec<&str> = col_def.splitn(2, ' ').collect();
-        if parts.len() != 2 {
-            return Err(Status::internal("Invalid column format"));
-        }
-        let name = parts[0].trim_matches('"').to_string();
-        let sql_type = parts[1].to_string();
-        columns.push((name, sql_type));
+    for col_def in stored_columns {
+        let col_name = col_def.name.trim().to_string();
+        let sql_type = map_field_type(&col_def.field_type)
+            .map_err(|e| Status::invalid_argument(format!(
+                "Invalid type for column '{}': {}",
+                col_name, e
+            )))?;
+        columns.push((col_name, sql_type));
     }
 
     // Build list of valid system columns (foreign keys and special columns)

server/tests/table_script/comprehensive_error_scenarios_tests.rs

@@ -3,6 +3,7 @@
 use crate::common::setup_isolated_db;
 use server::table_script::handlers::post_table_script::post_table_script;  // Fixed import
 use common::proto::komp_ac::table_script::PostTableScriptRequest;
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use rstest::*;
 use serde_json::json;
 use sqlx::PgPool;
@@ -12,15 +13,10 @@ async fn create_test_table(
     pool: &PgPool,
     schema_id: i64,
     table_name: &str,
-    columns: Vec<(&str, &str)>,
+    columns: Vec<ColumnDefinition>,
 ) -> i64 {
-    let column_definitions: Vec<String> = columns
-        .iter()
-        .map(|(name, type_def)| format!("\"{}\" {}", name, type_def))
-        .collect();
-
-    let columns_json = json!(column_definitions);
-    let indexes_json = json!([]);
+    let columns_json = serde_json::to_value(columns).unwrap();
+    let indexes_json = serde_json::json!([]);
 
     sqlx::query_scalar!(
         r#"INSERT INTO table_definitions (schema_id, table_name, columns, indexes)
@@ -115,22 +111,17 @@ async fn test_comprehensive_error_scenarios(
     let schema_id = get_default_schema_id(&pool).await;
 
     // Create comprehensive error test table
-    let columns = vec![
-        // Valid types
-        ("valid_numeric", "NUMERIC(10, 2)"),
-        ("valid_integer", "INTEGER"),
-
-        // Invalid for math operations
-        ("text_col", "TEXT"),
-        ("boolean_col", "BOOLEAN"),
-        ("bigint_col", "BIGINT"),
-        ("date_col", "DATE"),
-        ("timestamp_col", "TIMESTAMPTZ"),
-
-        // Invalid target types
-        ("bigint_target", "BIGINT"),
-        ("date_target", "DATE"),
-        ("timestamp_target", "TIMESTAMPTZ"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "valid_numeric".into(), field_type: "NUMERIC(10, 2)".into() },
+        ColumnDefinition { name: "valid_integer".into(), field_type: "INTEGER".into() },
+        ColumnDefinition { name: "text_col".into(), field_type: "TEXT".into() },
+        ColumnDefinition { name: "boolean_col".into(), field_type: "BOOLEAN".into() },
+        ColumnDefinition { name: "bigint_col".into(), field_type: "BIGINT".into() },
+        ColumnDefinition { name: "date_col".into(), field_type: "DATE".into() },
+        ColumnDefinition { name: "timestamp_col".into(), field_type: "TIMESTAMPTZ".into() },
+        ColumnDefinition { name: "bigint_target".into(), field_type: "BIGINT".into() },
+        ColumnDefinition { name: "date_target".into(), field_type: "DATE".into() },
+        ColumnDefinition { name: "timestamp_target".into(), field_type: "TIMESTAMPTZ".into() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "error_table", columns).await;
@@ -169,7 +160,9 @@ async fn test_malformed_script_scenarios(
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![("result", "NUMERIC(10, 2)")];
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "result".into(), field_type: "NUMERIC(10, 2)".into() }
+    ];
     let table_id = create_test_table(&pool, schema_id, "malformed_test", columns).await;
 
     let request = PostTableScriptRequest {
@@ -194,7 +187,9 @@ async fn test_advanced_validation_scenarios(
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![("result", "NUMERIC(10, 2)")];
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "result".into(), field_type: "NUMERIC(10, 2)".into() }
+    ];
     let table_id = create_test_table(&pool, schema_id, "advanced_test", columns).await;
 
     let request = PostTableScriptRequest {
@@ -236,16 +231,16 @@ async fn test_dependency_cycle_detection() {
     let schema_id = get_default_schema_id(&pool).await;
 
     // Create table_b first
-    let table_b_columns = vec![
-        ("value_b", "NUMERIC(10, 2)"),
-        ("result_b", "NUMERIC(10, 2)"),
+    let table_b_columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "value_b".into(), field_type: "NUMERIC(10, 2)".into() },
+        ColumnDefinition { name: "result_b".into(), field_type: "NUMERIC(10, 2)".into() },
     ];
     let table_b_id = create_test_table(&pool, schema_id, "table_b", table_b_columns).await;
 
     // Create table_a
-    let table_a_columns = vec![
-        ("value_a", "NUMERIC(10, 2)"),
-        ("result_a", "NUMERIC(10, 2)"),
+    let table_a_columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "value_a".into(), field_type: "NUMERIC(10, 2)".into() },
+        ColumnDefinition { name: "result_a".into(), field_type: "NUMERIC(10, 2)".into() },
     ];
     let table_a_id = create_test_table(&pool, schema_id, "table_a", table_a_columns).await;
 
@@ -305,7 +300,9 @@ async fn test_edge_case_identifiers(
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![("result", "NUMERIC(10, 2)")];
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "result".into(), field_type: "NUMERIC(10, 2)".into() }
+    ];
     let table_id = create_test_table(&pool, schema_id, "identifier_test", columns).await;
 
     // Test with edge case identifier in script
@@ -342,7 +339,9 @@ async fn test_sql_injection_prevention() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![("result", "NUMERIC(10, 2)")];
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "result".into(), field_type: "NUMERIC(10, 2)".into() }
+    ];
     let table_id = create_test_table(&pool, schema_id, "injection_test", columns).await;
 
     // Attempt SQL injection through script content
@@ -388,9 +387,9 @@ async fn test_performance_with_deeply_nested_expressions() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("x", "NUMERIC(15, 8)"),
-        ("performance_result", "NUMERIC(25, 12)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "x".into(), field_type: "NUMERIC(15, 8)".into() },
+        ColumnDefinition { name: "performance_result".into(), field_type: "NUMERIC(25, 12)".into() },
     ];
     let table_id = create_test_table(&pool, schema_id, "performance_test", columns).await;
 
@@ -437,11 +436,11 @@ async fn test_concurrent_script_creation() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("value", "NUMERIC(10, 2)"),
-        ("result1", "NUMERIC(10, 2)"),
-        ("result2", "NUMERIC(10, 2)"),
-        ("result3", "NUMERIC(10, 2)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "value".into(), field_type: "NUMERIC(10, 2)".into() },
+        ColumnDefinition { name: "result1".into(), field_type: "NUMERIC(10, 2)".into() },
+        ColumnDefinition { name: "result2".into(), field_type: "NUMERIC(10, 2)".into() },
+        ColumnDefinition { name: "result3".into(), field_type: "NUMERIC(10, 2)".into() },
     ];
     let table_id = create_test_table(&pool, schema_id, "concurrent_test", columns).await;
 
@@ -500,9 +499,10 @@ async fn test_error_message_localization_and_clarity() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("text_col", "TEXT"),
-        ("result", "NUMERIC(10, 2)"),
+
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "text_col".into(), field_type: "TEXT".into() },
+        ColumnDefinition { name: "result".into(), field_type: "NUMERIC(10, 2)".into() },
     ];
     let table_id = create_test_table(&pool, schema_id, "error_clarity_test", columns).await;
 

server/tests/table_script/mathematical_operations_tests.rs

@@ -3,6 +3,7 @@
 use crate::common::setup_isolated_db;
 use server::table_script::handlers::post_table_script::post_table_script;  // Fixed import
 use common::proto::komp_ac::table_script::PostTableScriptRequest;
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use rstest::*;
 use serde_json::json;
 use sqlx::PgPool;
@@ -12,15 +13,10 @@ async fn create_test_table(
     pool: &PgPool,
     schema_id: i64,
     table_name: &str,
-    columns: Vec<(&str, &str)>,
+    columns: Vec<ColumnDefinition>,
 ) -> i64 {
-    let column_definitions: Vec<String> = columns
-        .iter()
-        .map(|(name, type_def)| format!("\"{}\" {}", name, type_def))
-        .collect();
-
-    let columns_json = json!(column_definitions);
-    let indexes_json = json!([]);
+    let columns_json = serde_json::to_value(columns).unwrap();
+    let indexes_json = serde_json::json!([]);
 
     sqlx::query_scalar!(
         r#"INSERT INTO table_definitions (schema_id, table_name, columns, indexes)
@@ -97,7 +93,9 @@ async fn test_steel_decimal_literal_operations(
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![("result", "NUMERIC(30, 15)")];
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(30, 15)".to_string() }
+    ];
     let table_id = create_test_table(&pool, schema_id, "literal_test", columns).await;
 
     let script = format!(r#"({} "{}" "{}")"#, operation, value1, value2);
@@ -133,9 +131,9 @@ async fn test_steel_decimal_column_operations(
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("test_value", column_type),
-        ("result", "NUMERIC(30, 15)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "test_value".to_string(), field_type: column_type.to_string() },
+        ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(30, 15)".to_string() },
     ];
     let table_id = create_test_table(&pool, schema_id, "column_test", columns).await;
 
@@ -179,12 +177,12 @@ async fn test_complex_financial_calculation(
     let schema_id = get_default_schema_id(&pool).await;
 
     // Create a realistic financial calculation table
-    let columns = vec![
-        ("principal", "NUMERIC(16, 2)"),        // Principal amount
-        ("annual_rate", "NUMERIC(6, 5)"),       // Interest rate
-        ("years", "INTEGER"),                   // Time period
-        ("compounding_periods", "INTEGER"),     // Compounding frequency
-        ("compound_interest", "NUMERIC(20, 8)"), // Result
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "principal".to_string(), field_type: "NUMERIC(16, 2)".to_string() },        // Principal amount
+        ColumnDefinition { name: "annual_rate".to_string(), field_type: "NUMERIC(6, 5)".to_string() },       // Interest rate
+        ColumnDefinition { name: "years".to_string(), field_type: "INTEGER".to_string() },                   // Time period
+        ColumnDefinition { name: "compounding_periods".to_string(), field_type: "INTEGER".to_string() },     // Compounding frequency
+        ColumnDefinition { name: "compound_interest".to_string(), field_type: "NUMERIC(20, 8)".to_string() }, // Result
     ];
 
     let table_id = create_test_table(&pool, schema_id, "financial_calc", columns).await;
@@ -217,11 +215,11 @@ async fn test_scientific_precision_calculations() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("measurement_a", "NUMERIC(25, 15)"),
-        ("measurement_b", "NUMERIC(25, 15)"),
-        ("coefficient", "NUMERIC(10, 8)"),
-        ("scientific_result", "NUMERIC(30, 18)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "measurement_a".to_string(), field_type: "NUMERIC(25, 15)".to_string() },
+        ColumnDefinition { name: "measurement_b".to_string(), field_type: "NUMERIC(25, 15)".to_string() },
+        ColumnDefinition { name: "coefficient".to_string(), field_type: "NUMERIC(10, 8)".to_string() },
+        ColumnDefinition { name: "scientific_result".to_string(), field_type: "NUMERIC(30, 18)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "scientific_data", columns).await;
@@ -259,9 +257,9 @@ async fn test_precision_boundary_conditions(
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("boundary_value", numeric_type),
-        ("result", "NUMERIC(30, 15)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "boundary_value".to_string(), field_type: numeric_type.to_string() },
+        ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(30, 15)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "boundary_test", columns).await;
@@ -284,11 +282,11 @@ async fn test_mixed_integer_and_numeric_operations() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("integer_quantity", "INTEGER"),
-        ("numeric_price", "NUMERIC(10, 4)"),
-        ("numeric_tax_rate", "NUMERIC(5, 4)"),
-        ("total_with_tax", "NUMERIC(15, 4)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "integer_quantity".to_string(), field_type: "INTEGER".to_string() },
+        ColumnDefinition { name: "numeric_price".to_string(), field_type: "NUMERIC(10, 4)".to_string() },
+        ColumnDefinition { name: "numeric_tax_rate".to_string(), field_type: "NUMERIC(5, 4)".to_string() },
+        ColumnDefinition { name: "total_with_tax".to_string(), field_type: "NUMERIC(15, 4)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "mixed_types_calc", columns).await;
@@ -325,9 +323,9 @@ async fn test_mathematical_edge_cases(
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("test_value", "NUMERIC(15, 6)"),
-        ("result", "NUMERIC(20, 8)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "test_value".to_string(), field_type: "NUMERIC(15, 6)".to_string() },
+        ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(20, 8)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "edge_case_test", columns).await;
@@ -381,10 +379,10 @@ async fn test_comparison_operations_with_valid_types() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("value_a", "NUMERIC(10, 2)"),
-        ("value_b", "INTEGER"),
-        ("comparison_result", "BOOLEAN"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "value_a".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+        ColumnDefinition { name: "value_b".to_string(), field_type: "INTEGER".to_string() },
+        ColumnDefinition { name: "comparison_result".to_string(), field_type: "BOOLEAN".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "comparison_test", columns).await;
@@ -419,11 +417,11 @@ async fn test_nested_mathematical_expressions() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("x", "NUMERIC(15, 8)"),
-        ("y", "NUMERIC(15, 8)"),
-        ("z", "INTEGER"),
-        ("nested_result", "NUMERIC(25, 12)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "x".to_string(), field_type: "NUMERIC(15, 8)".to_string() },
+        ColumnDefinition { name: "y".to_string(), field_type: "NUMERIC(15, 8)".to_string() },
+        ColumnDefinition { name: "z".to_string(), field_type: "INTEGER".to_string() },
+        ColumnDefinition { name: "nested_result".to_string(), field_type: "NUMERIC(25, 12)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "nested_calc", columns).await;

server/tests/table_script/post_scripts_integration_tests.rs

@@ -3,6 +3,7 @@
 use crate::common::setup_isolated_db;
 use server::table_script::handlers::post_table_script::post_table_script;
 use common::proto::komp_ac::table_script::{PostTableScriptRequest, TableScriptResponse};
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use serde_json::json;
 use sqlx::PgPool;
 
@@ -26,14 +27,9 @@ impl TableScriptTestHelper {
         }
     }
 
-    pub async fn create_table_with_types(&self, table_name: &str, column_definitions: Vec<(&str, &str)>) -> i64 {
-        let columns: Vec<String> = column_definitions
-            .iter()
-            .map(|(name, type_def)| format!("\"{}\" {}", name, type_def))
-            .collect();
-
-        let columns_json = json!(columns);
-        let indexes_json = json!([]);
+    pub async fn create_table_with_types(&self, table_name: &str, column_definitions: Vec<ColumnDefinition>) -> i64 {
+        let columns_json = serde_json::to_value(column_definitions).unwrap();
+        let indexes_json = serde_json::json!([]);
 
         sqlx::query_scalar!(
             r#"INSERT INTO table_definitions (schema_id, table_name, columns, indexes)
@@ -73,24 +69,24 @@ mod integration_tests {
             "comprehensive_table",
             vec![
                 // Supported types for math operations
-                ("integer_col", "INTEGER"),
-                ("numeric_basic", "NUMERIC(10, 2)"),
-                ("numeric_high_precision", "NUMERIC(28, 15)"),
-                ("numeric_currency", "NUMERIC(14, 4)"),
+                ColumnDefinition { name: "integer_col".to_string(), field_type: "INTEGER".to_string() },
+                ColumnDefinition { name: "numeric_basic".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+                ColumnDefinition { name: "numeric_high_precision".to_string(), field_type: "NUMERIC(28, 15)".to_string() },
+                ColumnDefinition { name: "numeric_currency".to_string(), field_type: "NUMERIC(14, 4)".to_string() },
 
                 // Supported but not for math operations
-                ("text_col", "TEXT"),
-                ("boolean_col", "BOOLEAN"),
+                ColumnDefinition { name: "text_col".to_string(), field_type: "TEXT".to_string() },
+                ColumnDefinition { name: "boolean_col".to_string(), field_type: "BOOLEAN".to_string() },
 
                 // Prohibited types entirely
-                ("bigint_col", "BIGINT"),
-                ("date_col", "DATE"),
-                ("timestamp_col", "TIMESTAMPTZ"),
+                ColumnDefinition { name: "bigint_col".to_string(), field_type: "BIGINT".to_string() },
+                ColumnDefinition { name: "date_col".to_string(), field_type: "DATE".to_string() },
+                ColumnDefinition { name: "timestamp_col".to_string(), field_type: "TIMESTAMPTZ".to_string() },
 
                 // Result columns of various types
-                ("result_integer", "INTEGER"),
-                ("result_numeric", "NUMERIC(15, 5)"),
-                ("result_text", "TEXT"),
+                ColumnDefinition { name: "result_integer".to_string(), field_type: "INTEGER".to_string() },
+                ColumnDefinition { name: "result_numeric".to_string(), field_type: "NUMERIC(15, 5)".to_string() },
+                ColumnDefinition { name: "result_text".to_string(), field_type: "TEXT".to_string() },
             ]
         ).await;
 
@@ -150,13 +146,13 @@ mod integration_tests {
         let table_id = helper.create_table_with_types(
             "precision_table",
             vec![
-                ("low_precision", "NUMERIC(5, 2)"),      // e.g., 999.99
-                ("medium_precision", "NUMERIC(10, 4)"),  // e.g., 999999.9999
-                ("high_precision", "NUMERIC(28, 15)"),   // Maximum PostgreSQL precision
-                ("currency", "NUMERIC(14, 4)"),          // Standard currency precision
-                ("percentage", "NUMERIC(5, 4)"),         // e.g., 0.9999 (99.99%)
-                ("integer_val", "INTEGER"),
-                ("result", "NUMERIC(30, 15)"),
+                ColumnDefinition { name: "low_precision".to_string(), field_type: "NUMERIC(5, 2)".to_string() },      // e.g., 999.99
+                ColumnDefinition { name: "medium_precision".to_string(), field_type: "NUMERIC(10, 4)".to_string() },  // e.g., 999999.9999
+                ColumnDefinition { name: "high_precision".to_string(), field_type: "NUMERIC(28, 15)".to_string() },   // Maximum PostgreSQL precision
+                ColumnDefinition { name: "currency".to_string(), field_type: "NUMERIC(14, 4)".to_string() },          // Standard currency precision
+                ColumnDefinition { name: "percentage".to_string(), field_type: "NUMERIC(5, 4)".to_string() },         // e.g., 0.9999 (99.99%)
+                ColumnDefinition { name: "integer_val".to_string(), field_type: "INTEGER".to_string() },
+                ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(30, 15)".to_string() },
             ]
         ).await;
 
@@ -202,12 +198,12 @@ mod integration_tests {
         let table_id = helper.create_table_with_types(
             "financial_instruments",
             vec![
-                ("principal", "NUMERIC(16, 2)"),      // Principal amount
-                ("annual_rate", "NUMERIC(6, 5)"),     // Interest rate (e.g., 0.05250)
-                ("years", "INTEGER"),                 // Time period
-                ("compounding_periods", "INTEGER"),   // Compounding frequency
-                ("fees", "NUMERIC(10, 2)"),          // Transaction fees
-                ("compound_interest", "NUMERIC(20, 8)"), // Result column
+                ColumnDefinition { name: "principal".to_string(), field_type: "NUMERIC(16, 2)".to_string() },      // Principal amount
+                ColumnDefinition { name: "annual_rate".to_string(), field_type: "NUMERIC(6, 5)".to_string() },     // Interest rate (e.g., 0.05250)
+                ColumnDefinition { name: "years".to_string(), field_type: "INTEGER".to_string() },                 // Time period
+                ColumnDefinition { name: "compounding_periods".to_string(), field_type: "INTEGER".to_string() },   // Compounding frequency
+                ColumnDefinition { name: "fees".to_string(), field_type: "NUMERIC(10, 2)".to_string() },          // Transaction fees
+                ColumnDefinition { name: "compound_interest".to_string(), field_type: "NUMERIC(20, 8)".to_string() }, // Result column
             ]
         ).await;
 
@@ -237,9 +233,9 @@ mod integration_tests {
         let table_id = helper.create_table_with_types(
             "scientific_data",
             vec![
-                ("large_number", "NUMERIC(30, 10)"),
-                ("small_number", "NUMERIC(30, 20)"),
-                ("result", "NUMERIC(35, 25)"),
+                ColumnDefinition { name: "large_number".to_string(), field_type: "NUMERIC(30, 10)".to_string() },
+                ColumnDefinition { name: "small_number".to_string(), field_type: "NUMERIC(30, 20)".to_string() },
+                ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(35, 25)".to_string() },
             ]
         ).await;
 
@@ -265,8 +261,8 @@ mod integration_tests {
         let table_a_id = helper.create_table_with_types(
             "table_a",
             vec![
-                ("value_a", "NUMERIC(10, 2)"),
-                ("result_a", "NUMERIC(10, 2)"),
+                ColumnDefinition { name: "value_a".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+                ColumnDefinition { name: "result_a".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
             ]
         ).await;
         println!("Created table_a with ID: {}", table_a_id);
@@ -274,8 +270,8 @@ mod integration_tests {
         let table_b_id = helper.create_table_with_types(
             "table_b",
             vec![
-                ("value_b", "NUMERIC(10, 2)"),
-                ("result_b", "NUMERIC(10, 2)"),
+                ColumnDefinition { name: "value_b".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+                ColumnDefinition { name: "result_b".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
             ]
         ).await;
         println!("Created table_b with ID: {}", table_b_id);
@@ -354,10 +350,10 @@ mod integration_tests {
         let table_id = helper.create_table_with_types(
             "error_test_table",
             vec![
-                ("text_field", "TEXT"),
-                ("numeric_field", "NUMERIC(10, 2)"),
-                ("boolean_field", "BOOLEAN"),
-                ("bigint_field", "BIGINT"),
+                ColumnDefinition { name: "text_field".to_string(), field_type: "TEXT".to_string() },
+                ColumnDefinition { name: "numeric_field".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+                ColumnDefinition { name: "boolean_field".to_string(), field_type: "BOOLEAN".to_string() },
+                ColumnDefinition { name: "bigint_field".to_string(), field_type: "BIGINT".to_string() },
             ]
         ).await;
 
@@ -417,11 +413,11 @@ mod integration_tests {
         let table_id = helper.create_table_with_types(
             "performance_table",
             vec![
-                ("x", "NUMERIC(15, 8)"),
-                ("y", "NUMERIC(15, 8)"),
-                ("z", "NUMERIC(15, 8)"),
-                ("w", "NUMERIC(15, 8)"),
-                ("complex_result", "NUMERIC(25, 12)"),
+                ColumnDefinition { name: "x".to_string(), field_type: "NUMERIC(15, 8)".to_string() },
+                ColumnDefinition { name: "y".to_string(), field_type: "NUMERIC(15, 8)".to_string() },
+                ColumnDefinition { name: "z".to_string(), field_type: "NUMERIC(15, 8)".to_string() },
+                ColumnDefinition { name: "w".to_string(), field_type: "NUMERIC(15, 8)".to_string() },
+                ColumnDefinition { name: "complex_result".to_string(), field_type: "NUMERIC(25, 12)".to_string() },
             ]
         ).await;
 
@@ -456,11 +452,11 @@ mod integration_tests {
         let table_id = helper.create_table_with_types(
             "boundary_table",
             vec![
-                ("min_numeric", "NUMERIC(1, 0)"),      // Minimum: single digit, no decimal
-                ("max_numeric", "NUMERIC(1000, 999)"),  // Maximum PostgreSQL allows
-                ("zero_scale", "NUMERIC(10, 0)"),      // Integer-like numeric
-                ("max_scale", "NUMERIC(28, 28)"),      // Maximum scale
-                ("result", "NUMERIC(1000, 999)"),
+                ColumnDefinition { name: "min_numeric".to_string(), field_type: "NUMERIC(1, 0)".to_string() },      // Minimum: single digit, no decimal
+                ColumnDefinition { name: "max_numeric".to_string(), field_type: "NUMERIC(1000, 999)".to_string() },  // Maximum PostgreSQL allows
+                ColumnDefinition { name: "zero_scale".to_string(), field_type: "NUMERIC(10, 0)".to_string() },      // Integer-like numeric
+                ColumnDefinition { name: "max_scale".to_string(), field_type: "NUMERIC(28, 28)".to_string() },      // Maximum scale
+                ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(1000, 999)".to_string() },
             ]
         ).await;
 
@@ -495,10 +491,10 @@ mod steel_decimal_integration_tests {
         let table_id = helper.create_table_with_types(
             "test_execution_table",
             vec![
-                ("amount", "NUMERIC(10, 2)"),
-                ("quantity", "INTEGER"),
-                ("tax_rate", "NUMERIC(5, 4)"),
-                ("result", "NUMERIC(15, 4)"), // Add a result column
+                ColumnDefinition { name: "amount".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+                ColumnDefinition { name: "quantity".to_string(), field_type: "INTEGER".to_string() },
+                ColumnDefinition { name: "tax_rate".to_string(), field_type: "NUMERIC(5, 4)".to_string() },
+                ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(15, 4)".to_string() }, // Add a result column
             ]
         ).await;
         println!("Created test table with ID: {}", table_id);
@@ -575,9 +571,9 @@ mod steel_decimal_integration_tests {
         let table_id = helper.create_table_with_types(
             "precision_test_table",
             vec![
-                ("precise_value", "NUMERIC(20, 12)"),
-                ("multiplier", "NUMERIC(20, 12)"),
-                ("result", "NUMERIC(25, 15)"), // Add result column
+                ColumnDefinition { name: "precise_value".to_string(), field_type: "NUMERIC(20, 12)".to_string() },
+                ColumnDefinition { name: "multiplier".to_string(), field_type: "NUMERIC(20, 12)".to_string() },
+                ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(25, 15)".to_string() }, // Add result column
             ]
         ).await;
         println!("Created precision test table with ID: {}", table_id);

server/tests/table_script/prohibited_types_test.rs

@@ -3,6 +3,7 @@
 use crate::common::setup_isolated_db;
 use server::table_script::handlers::post_table_script::post_table_script;
 use common::proto::komp_ac::table_script::PostTableScriptRequest;
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use serde_json::json;
 use sqlx::PgPool;
 
@@ -11,15 +12,10 @@ async fn create_test_table(
     pool: &PgPool,
     schema_id: i64,
     table_name: &str,
-    columns: Vec<(&str, &str)>,
+    columns: Vec<ColumnDefinition>,
 ) -> i64 {
-    let column_definitions: Vec<String> = columns
-        .iter()
-        .map(|(name, type_def)| format!("\"{}\" {}", name, type_def))
-        .collect();
-
-    let columns_json = json!(column_definitions);
-    let indexes_json = json!([]);
+    let columns_json = serde_json::to_value(columns).unwrap();
+    let indexes_json = serde_json::json!([]);
 
     sqlx::query_scalar!(
         r#"INSERT INTO table_definitions (schema_id, table_name, columns, indexes)
@@ -67,7 +63,10 @@ async fn test_reject_bigint_target_column() {
         &pool,
         schema_id,
         "bigint_table",
-        vec![("name", "TEXT"), ("big_number", "BIGINT")]
+        vec![
+            ColumnDefinition { name: "name".to_string(), field_type: "TEXT".to_string() },
+            ColumnDefinition { name: "big_number".to_string(), field_type: "BIGINT".to_string() }
+        ]
     ).await;
 
     let request = PostTableScriptRequest {
@@ -99,7 +98,10 @@ async fn test_reject_date_target_column() {
         &pool,
         schema_id,
         "date_table",
-        vec![("name", "TEXT"), ("event_date", "DATE")]
+        vec![
+            ColumnDefinition { name: "name".to_string(), field_type: "TEXT".to_string() },
+            ColumnDefinition { name: "event_date".to_string(), field_type: "DATE".to_string() }
+        ]
     ).await;
 
     let request = PostTableScriptRequest {
@@ -131,7 +133,10 @@ async fn test_reject_timestamptz_target_column() {
         &pool,
         schema_id,
         "timestamp_table",
-        vec![("name", "TEXT"), ("created_time", "TIMESTAMPTZ")]
+        vec![
+            ColumnDefinition { name: "name".to_string(), field_type: "TEXT".to_string() },
+            ColumnDefinition { name: "created_time".to_string(), field_type: "TIMESTAMPTZ".to_string() }
+        ]
     ).await;
 
     let request = PostTableScriptRequest {
@@ -164,9 +169,9 @@ async fn test_reject_text_in_mathematical_operations() {
         schema_id,
         "text_math_table",
         vec![
-            ("description", "TEXT"),
-            ("amount", "NUMERIC(10, 2)"),
-            ("result", "NUMERIC(10, 2)")
+            ColumnDefinition { name: "description".to_string(), field_type: "TEXT".to_string() },
+            ColumnDefinition { name: "amount".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+            ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(10, 2)".to_string() }
         ]
     ).await;
 
@@ -202,9 +207,9 @@ async fn test_reject_boolean_in_mathematical_operations() {
         schema_id,
         "boolean_math_table",
         vec![
-            ("is_active", "BOOLEAN"),
-            ("amount", "NUMERIC(10, 2)"),
-            ("result", "NUMERIC(10, 2)")
+            ColumnDefinition { name: "is_active".to_string(), field_type: "BOOLEAN".to_string() },
+            ColumnDefinition { name: "amount".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+            ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(10, 2)".to_string() }
         ]
     ).await;
 
@@ -240,8 +245,8 @@ async fn test_reject_bigint_in_mathematical_operations() {
         schema_id,
         "bigint_math_table",
         vec![
-            ("big_value", "BIGINT"),
-            ("result", "NUMERIC(10, 2)")
+            ColumnDefinition { name: "big_value".to_string(), field_type: "BIGINT".to_string() },
+            ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(10, 2)".to_string() }
         ]
     ).await;
 
@@ -277,10 +282,10 @@ async fn test_allow_valid_script_with_allowed_types() {
         schema_id,
         "allowed_types_table",
         vec![
-            ("name", "TEXT"),
-            ("count", "INTEGER"),
-            ("amount", "NUMERIC(10, 2)"),
-            ("computed_value", "TEXT")
+            ColumnDefinition { name: "name".to_string(), field_type: "TEXT".to_string() },
+            ColumnDefinition { name: "count".to_string(), field_type: "INTEGER".to_string() },
+            ColumnDefinition { name: "amount".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+            ColumnDefinition { name: "computed_value".to_string(), field_type: "TEXT".to_string() }
         ]
     ).await;
 
@@ -312,9 +317,9 @@ async fn test_allow_integer_and_numeric_in_math_operations() {
         schema_id,
         "math_allowed_table",
         vec![
-            ("quantity", "INTEGER"),
-            ("price", "NUMERIC(10, 2)"),
-            ("total", "NUMERIC(12, 2)")
+            ColumnDefinition { name: "quantity".to_string(), field_type: "INTEGER".to_string() },
+            ColumnDefinition { name: "price".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+            ColumnDefinition { name: "total".to_string(), field_type: "NUMERIC(12, 2)".to_string() }
         ]
     ).await;
 
@@ -328,7 +333,7 @@ async fn test_allow_integer_and_numeric_in_math_operations() {
     )
     .fetch_one(&pool)
     .await;
-    
+
     println!("Table verification: {:?}", table_check);
 
     let request = PostTableScriptRequest {
@@ -341,7 +346,7 @@ async fn test_allow_integer_and_numeric_in_math_operations() {
 
     println!("About to call post_table_script");
     let result = post_table_script(&pool, request).await;
-    
+
     // SHOW THE ACTUAL ERROR
     if let Err(e) = &result {
         println!("ERROR: {}", e);
@@ -363,14 +368,19 @@ async fn test_script_without_table_links_should_fail() {
         &pool,
         schema_id,
         "table_a",
-        vec![("value_a", "INTEGER"), ("result", "INTEGER")]
+        vec![
+            ColumnDefinition { name: "value_a".to_string(), field_type: "INTEGER".to_string() },
+            ColumnDefinition { name: "result".to_string(), field_type: "INTEGER".to_string() }
+        ]
     ).await;
 
     let _table_b_id = create_test_table(
         &pool,
         schema_id,
-        "table_b", 
-        vec![("value_b", "INTEGER")]
+        "table_b",
+        vec![
+            ColumnDefinition { name: "value_b".to_string(), field_type: "INTEGER".to_string() }
+        ]
     ).await;
 
     // DON'T create a link between the tables
@@ -404,14 +414,19 @@ async fn test_script_with_table_links_should_succeed() {
         &pool,
         schema_id,
         "linked_table_a",
-        vec![("value_a", "INTEGER"), ("result", "INTEGER")]
+        vec![
+            ColumnDefinition { name: "value_a".to_string(), field_type: "INTEGER".to_string() },
+            ColumnDefinition { name: "result".to_string(), field_type: "INTEGER".to_string() }
+        ]
     ).await;
 
     let table_b_id = create_test_table(
         &pool,
         schema_id,
         "linked_table_b",
-        vec![("value_b", "INTEGER")]
+        vec![
+            ColumnDefinition { name: "value_b".to_string(), field_type: "INTEGER".to_string() }
+        ]
     ).await;
 
     // Create a link between the tables (table_a can access table_b)

server/tests/table_script/type_safety_comprehensive_tests.rs

@@ -3,6 +3,7 @@
 use crate::common::setup_isolated_db;
 use server::table_script::handlers::post_table_script::post_table_script;
 use common::proto::komp_ac::table_script::PostTableScriptRequest;
+use common::proto::komp_ac::table_definition::ColumnDefinition;
 use rstest::*;
 use serde_json::json;
 use sqlx::PgPool;
@@ -76,15 +77,10 @@ async fn create_test_table(
     pool: &PgPool,
     schema_id: i64,
     table_name: &str,
-    columns: Vec<(&str, &str)>,
+    columns: Vec<ColumnDefinition>,
 ) -> i64 {
-    let column_definitions: Vec<String> = columns
-        .iter()
-        .map(|(name, type_def)| format!("\"{}\" {}", name, type_def))
-        .collect();
-
-    let columns_json = json!(column_definitions);
-    let indexes_json = json!([]);
+    let columns_json = serde_json::to_value(columns).unwrap();
+    let indexes_json = serde_json::json!([]);
 
     sqlx::query_scalar!(
         r#"INSERT INTO table_definitions (schema_id, table_name, columns, indexes)
@@ -123,8 +119,17 @@ async fn test_allowed_types_in_math_operations(
     let schema_id = get_default_schema_id(&pool).await;
 
     // Create table with all allowed mathematical types plus a result column
-    let mut columns = allowed_math_types.clone();
-    columns.push(("result", "NUMERIC(30, 15)"));
+    let mut columns: Vec<ColumnDefinition> = allowed_math_types
+        .iter()
+        .map(|(name, field_type)| ColumnDefinition {
+            name: name.to_string(),
+            field_type: field_type.to_string(),
+        })
+        .collect();
+    columns.push(ColumnDefinition {
+        name: "result".to_string(),
+        field_type: "NUMERIC(30, 15)".to_string(),
+    });
 
     let table_id = create_test_table(&pool, schema_id, "math_test_table", columns).await;
 
@@ -172,8 +177,17 @@ async fn test_prohibited_types_in_math_operations(
     let schema_id = get_default_schema_id(&pool).await;
 
     // Create table with prohibited types plus a valid result column
-    let mut columns = prohibited_math_types.clone();
-    columns.push(("result", "NUMERIC(15, 6)"));
+    let mut columns: Vec<ColumnDefinition> = prohibited_math_types
+        .iter()
+        .map(|(name, field_type)| ColumnDefinition {
+            name: name.to_string(),
+            field_type: field_type.to_string(),
+        })
+        .collect();
+    columns.push(ColumnDefinition {
+        name: "result".to_string(),
+        field_type: "NUMERIC(15, 6)".to_string(),
+    });
 
     let table_id = create_test_table(&pool, schema_id, "prohibited_math_table", columns).await;
 
@@ -225,8 +239,17 @@ async fn test_prohibited_target_column_types(
     let schema_id = get_default_schema_id(&pool).await;
 
     // Create table with prohibited target types plus some valid source columns
-    let mut columns = prohibited_target_types.clone();
-    columns.push(("amount", "NUMERIC(10, 2)"));
+    let mut columns: Vec<ColumnDefinition> = prohibited_target_types
+        .iter()
+        .map(|(name, field_type)| ColumnDefinition {
+            name: name.to_string(),
+            field_type: field_type.to_string(),
+        })
+        .collect();
+    columns.push(ColumnDefinition {
+        name: "amount".to_string(),
+        field_type: "NUMERIC(10, 2)".to_string(),
+    });
 
     let table_id = create_test_table(&pool, schema_id, "prohibited_target_table", columns).await;
 
@@ -245,7 +268,7 @@ async fn test_prohibited_target_column_types(
 
     let error_message = result.unwrap_err().to_string();
     assert!(
-        error_message.to_lowercase().contains("cannot create script") || 
+        error_message.to_lowercase().contains("cannot create script") ||
         error_message.contains("prohibited type"),
         "Error should mention prohibited type: {}",
         error_message
@@ -261,7 +284,12 @@ async fn test_system_column_restrictions(#[case] target_column: &str, #[case] de
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![("amount", "NUMERIC(10, 2)")];
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition {
+            name: "amount".to_string(),
+            field_type: "NUMERIC(10, 2)".to_string(),
+        }
+    ];
     let table_id = create_test_table(&pool, schema_id, "system_test_table", columns).await;
 
     let script = r#"(+ "10" "20")"#;
@@ -290,22 +318,22 @@ async fn test_comprehensive_type_matrix() {
     let schema_id = get_default_schema_id(&pool).await;
 
     // Create comprehensive table with all type categories
-    let all_columns = vec![
+    let all_columns: Vec<ColumnDefinition> = vec![
         // Allowed math types
-        ("integer_col", "INTEGER"),
-        ("numeric_col", "NUMERIC(10, 2)"),
-        ("high_precision", "NUMERIC(28, 15)"),
-        
+        ColumnDefinition { name: "integer_col".to_string(), field_type: "INTEGER".to_string() },
+        ColumnDefinition { name: "numeric_col".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+        ColumnDefinition { name: "high_precision".to_string(), field_type: "NUMERIC(28, 15)".to_string() },
+
         // Prohibited math types
-        ("text_col", "TEXT"),
-        ("boolean_col", "BOOLEAN"),
-        ("bigint_col", "BIGINT"),
-        ("date_col", "DATE"),
-        ("timestamp_col", "TIMESTAMPTZ"),
-        
+        ColumnDefinition { name: "text_col".to_string(), field_type: "TEXT".to_string() },
+        ColumnDefinition { name: "boolean_col".to_string(), field_type: "BOOLEAN".to_string() },
+        ColumnDefinition { name: "bigint_col".to_string(), field_type: "BIGINT".to_string() },
+        ColumnDefinition { name: "date_col".to_string(), field_type: "DATE".to_string() },
+        ColumnDefinition { name: "timestamp_col".to_string(), field_type: "TIMESTAMPTZ".to_string() },
+
         // Result columns
-        ("result_numeric", "NUMERIC(20, 8)"),
-        ("result_text", "TEXT"),
+        ColumnDefinition { name: "result_numeric".to_string(), field_type: "NUMERIC(20, 8)".to_string() },
+        ColumnDefinition { name: "result_text".to_string(), field_type: "TEXT".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "comprehensive_table", all_columns).await;
@@ -316,7 +344,7 @@ async fn test_comprehensive_type_matrix() {
         ("integer_col", "+", "result_numeric", true),
         ("numeric_col", "*", "result_numeric", true),
         ("high_precision", "/", "result_numeric", true),
-        
+
         // Invalid source types in math
         ("text_col", "+", "result_numeric", false),
         ("boolean_col", "*", "result_numeric", false),
@@ -361,20 +389,20 @@ async fn test_complex_mathematical_expressions() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("principal", "NUMERIC(16, 2)"),
-        ("rate", "NUMERIC(6, 5)"),
-        ("years", "INTEGER"),
-        ("compound_result", "NUMERIC(20, 8)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "principal".to_string(), field_type: "NUMERIC(16, 2)".to_string() },
+        ColumnDefinition { name: "rate".to_string(), field_type: "NUMERIC(6, 5)".to_string() },
+        ColumnDefinition { name: "years".to_string(), field_type: "INTEGER".to_string() },
+        ColumnDefinition { name: "compound_result".to_string(), field_type: "NUMERIC(20, 8)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "financial_table", columns).await;
 
     // Complex compound interest calculation - all using allowed types
     let complex_script = r#"
-        (* 
+        (*
             (steel_get_column "financial_table" "principal")
-            (pow 
+            (pow
                 (+ "1" (steel_get_column "financial_table" "rate"))
                 (steel_get_column "financial_table" "years")))
     "#;
@@ -395,9 +423,9 @@ async fn test_nonexistent_column_reference() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("amount", "NUMERIC(10, 2)"),
-        ("result", "NUMERIC(10, 2)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "amount".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+        ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "simple_table", columns).await;
@@ -427,9 +455,9 @@ async fn test_nonexistent_table_reference() {
     let pool = setup_isolated_db().await;
     let schema_id = get_default_schema_id(&pool).await;
 
-    let columns = vec![
-        ("amount", "NUMERIC(10, 2)"),
-        ("result", "NUMERIC(10, 2)"),
+    let columns: Vec<ColumnDefinition> = vec![
+        ColumnDefinition { name: "amount".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
+        ColumnDefinition { name: "result".to_string(), field_type: "NUMERIC(10, 2)".to_string() },
     ];
 
     let table_id = create_test_table(&pool, schema_id, "existing_table", columns).await;

server/tests/tables_data/get/get_table_data_test.rs

@@ -67,10 +67,10 @@ async fn table_definition(#[future] schema: (PgPool, String, i64)) -> (PgPool, S
 
     // Define columns and indexes for the table
     let columns = json!([
-        "\"name\" TEXT",
-        "\"age\" INTEGER",
-        "\"email\" TEXT",
-        "\"is_active\" BOOLEAN"
+        { "name": "name", "field_type": "text" },
+        { "name": "age", "field_type": "integer" },
+        { "name": "email", "field_type": "text" },
+        { "name": "is_active", "field_type": "boolean" }
     ]);
     let indexes = json!([]);
 

server/tests/tables_data/put/put_table_data_steel_decimal_test.rs

@@ -91,7 +91,7 @@ async fn create_initial_record(
     // Set different initial values based on the test case to satisfy validation scripts
     match (profile_name, table_name) {
         ("test_put_complex", "order") => {
-            // For complex formula: (+ (* @price @quantity) (* (* @price @quantity) 0.08))
+            // For complex formula: (+ (* $price $quantity) (* (* $price $quantity) 0.08))
             // With price=10.00, quantity=1: (10*1) + (10*1*0.08) = 10 + 0.8 = 10.8
             data.insert("price".to_string(), ProtoValue { kind: Some(Kind::StringValue("10.00".to_string())) });
             data.insert("quantity".to_string(), ProtoValue { kind: Some(Kind::NumberValue(1.0)) });
@@ -99,7 +99,7 @@ async fn create_initial_record(
             data.insert("percentage".to_string(), ProtoValue { kind: Some(Kind::StringValue("100.00".to_string())) });
         },
         ("test_put_division", "calculation") => {
-            // For division: (/ @total @price) 
+            // For division: (/ $total $price)
             // With total=10.00, price=10.00: 10/10 = 1
             data.insert("price".to_string(), ProtoValue { kind: Some(Kind::StringValue("10.00".to_string())) });
             data.insert("quantity".to_string(), ProtoValue { kind: Some(Kind::NumberValue(1.0)) });
@@ -142,7 +142,7 @@ async fn test_put_basic_arithmetic_validation_success(pool: PgPool) {
     let script_request = PostTableScriptRequest {
         table_definition_id: table_def_id,
         target_column: "total".to_string(),
-        script: "(* @price @quantity)".to_string(),
+        script: "(* $price $quantity)".to_string(),
         description: "Total = Price × Quantity".to_string(),
     };
     post_table_script(&pool, script_request).await.unwrap();
@@ -180,7 +180,7 @@ async fn test_put_basic_arithmetic_validation_failure(pool: PgPool) {
     let script_request = PostTableScriptRequest {
         table_definition_id: table_def_id,
         target_column: "total".to_string(),
-        script: "(* @price @quantity)".to_string(),
+        script: "(* $price $quantity)".to_string(),
         description: "Total = Price × Quantity".to_string(),
     };
     post_table_script(&pool, script_request).await.unwrap();
@@ -224,7 +224,7 @@ async fn test_put_complex_formula_validation(pool: PgPool) {
     let script_request = PostTableScriptRequest {
         table_definition_id: table_def_id,
         target_column: "total".to_string(),
-        script: "(+ (* @price @quantity) (* (* @price @quantity) 0.08))".to_string(),
+        script: "(+ (* $price $quantity) (* (* $price $quantity) 0.08))".to_string(),
         description: "Total with 8% tax".to_string(),
     };
     post_table_script(&pool, script_request).await.unwrap();
@@ -261,7 +261,7 @@ async fn test_put_division_with_precision(pool: PgPool) {
     let script_request = PostTableScriptRequest {
         table_definition_id: table_def_id,
         target_column: "percentage".to_string(),
-        script: "(/ @total @price)".to_string(),
+        script: "(/ $total $price)".to_string(),
         description: "Percentage = Total / Price".to_string(),
     };
     post_table_script(&pool, script_request).await.unwrap();
@@ -326,7 +326,7 @@ async fn test_put_advanced_math_functions(pool: PgPool) {
     let sqrt_script = PostTableScriptRequest {
         table_definition_id: table_row.id,
         target_column: "square_root".to_string(),
-        script: "(sqrt @input)".to_string(),
+        script: "(sqrt $input)".to_string(),
         description: "Square root validation".to_string(),
     };
     post_table_script(&pool, sqrt_script).await.unwrap();
@@ -334,7 +334,7 @@ async fn test_put_advanced_math_functions(pool: PgPool) {
     let power_script = PostTableScriptRequest {
         table_definition_id: table_row.id,
         target_column: "power_result".to_string(),
-        script: "(^ @input 2.0)".to_string(),
+        script: "(^ $input 2.0)".to_string(),
         description: "Power function validation".to_string(),
     };
     post_table_script(&pool, power_script).await.unwrap();
@@ -389,7 +389,7 @@ async fn test_put_financial_calculations(pool: PgPool) {
     let compound_script = PostTableScriptRequest {
         table_definition_id: table_row.id,
         target_column: "compound_result".to_string(),
-        script: "(* @principal (^ (+ 1.0 @rate) @time))".to_string(),
+        script: "(* $principal (^ (+ 1.0 $rate) $time))".to_string(),
         description: "Compound interest calculation".to_string(),
     };
     post_table_script(&pool, compound_script).await.unwrap();
@@ -397,7 +397,7 @@ async fn test_put_financial_calculations(pool: PgPool) {
     let percentage_script = PostTableScriptRequest {
         table_definition_id: table_row.id,
         target_column: "percentage_result".to_string(),
-        script: "(* @principal @rate)".to_string(),
+        script: "(* $principal $rate)".to_string(),
         description: "Percentage calculation".to_string(),
     };
     post_table_script(&pool, percentage_script).await.unwrap();
@@ -441,15 +441,13 @@ async fn test_put_partial_update_with_validation(pool: PgPool) {
     let script_request = PostTableScriptRequest {
         table_definition_id: table_def_id,
         target_column: "total".to_string(),
-        script: "(* @price @quantity)".to_string(),
+        script: r#"( * (get-var "price") (get-var "quantity") )"#.to_string(),
         description: "Total = Price × Quantity".to_string(),
     };
     post_table_script(&pool, script_request).await.unwrap();
 
     let record_id = create_initial_record(&pool, "test_put_partial", "invoice", &indexer_tx).await;
 
-    // Partial update: only update quantity. The script detects this would change total
-    // from 10.00 to 50.00 and requires the user to include 'total' in the update.
     let mut update_data = HashMap::new();
     update_data.insert("quantity".to_string(), ProtoValue {
         kind: Some(Kind::NumberValue(5.0)),
@@ -462,16 +460,16 @@ async fn test_put_partial_update_with_validation(pool: PgPool) {
         data: update_data,
     };
 
-    // This should fail because script would change total value
+    // This should fail because script would change total value (Case B: implicit change detection)
     let result = put_table_data(&pool, put_request, &indexer_tx).await;
     assert!(result.is_err());
     let error = result.unwrap_err();
     assert_eq!(error.code(), tonic::Code::FailedPrecondition);
-    assert!(error.message().contains("Script for column 'total' was triggered"));
-    assert!(error.message().contains("from '10.00' to '50.00'"));
+    let msg = error.message();
+    assert!(msg.contains("Script for column 'total' was triggered"));
+    assert!(msg.contains("from '10.00' to '50.00'"));
+    assert!(msg.contains("include 'total' in your update request"));  // Full change detection msg
 
-    // Now, test a partial update that SHOULD fail validation.
-    // We update quantity and provide an incorrect total.
     let mut failing_update_data = HashMap::new();
     failing_update_data.insert("quantity".to_string(), ProtoValue {
         kind: Some(Kind::NumberValue(3.0)),
@@ -491,8 +489,9 @@ async fn test_put_partial_update_with_validation(pool: PgPool) {
     assert!(result.is_err());
     let error = result.unwrap_err();
     assert_eq!(error.code(), tonic::Code::InvalidArgument);
-    assert!(error.message().contains("Script calculated '30.00'"));
-    assert!(error.message().contains("but user provided '99.99'"));
+    let msg = error.message();
+    assert!(msg.contains("Script calculated '30.00'"));
+    assert!(msg.contains("but user provided '99.99'"));
 }
 
 #[sqlx::test]
@@ -553,7 +552,7 @@ async fn test_put_steel_script_error_handling(pool: PgPool) {
     let script_request = PostTableScriptRequest {
         table_definition_id: table_def_id,
         target_column: "total".to_string(),
-        script: "(/ @price 0.0)".to_string(),
+        script: "(/ $price 0.0)".to_string(),
         description: "Error test".to_string(),
     };
     post_table_script(&pool, script_request).await.unwrap();
@@ -623,7 +622,7 @@ async fn test_decimal_precision_behavior(pool: PgPool) {
     let script_request = PostTableScriptRequest {
         table_definition_id: table_row.id,
         target_column: "result".to_string(),
-        script: "(/ @dividend @divisor)".to_string(),
+        script: "(/ $dividend $divisor)".to_string(),
         description: "Division test for precision".to_string(),
     };
     post_table_script(&pool, script_request).await.unwrap();
@@ -816,7 +815,7 @@ async fn test_put_complex_formula_validation_via_handlers(pool: PgPool) {
         "test_put_complex_handlers",
         "order",
         "total",
-        "(+ (* @price @quantity) (* (* @price @quantity) 0.08))", // Total with 8% tax
+        "(+ (* $price $quantity) (* (* $price $quantity) 0.08))", // Total with 8% tax
     )
     .await
     .expect("Failed to add validation script");
@@ -891,7 +890,7 @@ async fn test_put_basic_arithmetic_validation_via_handlers(pool: PgPool) {
         "test_put_arithmetic_handlers",
         "invoice",
         "total",
-        "(* @price @quantity)", // Simple: Total = Price × Quantity
+        "(* $price $quantity)", // Simple: Total = Price × Quantity
     )
     .await
     .expect("Failed to add validation script");
@@ -955,7 +954,7 @@ async fn test_put_arithmetic_validation_failure_via_handlers(pool: PgPool) {
         "test_put_arithmetic_fail_handlers",
         "invoice",
         "total",
-        "(* @price @quantity)",
+        "(* $price $quantity)",
     )
     .await
     .expect("Failed to add validation script");