awslabs.s3-tables-mcp-server 0.0.1__tar.gz → 0.0.3__tar.gz

awslabs_s3_tables_mcp_server-0.0.3/CONTEXT.md

@@ -0,0 +1,110 @@
+# AWS S3 Tables MCP (Model Context Protocol) Server
+
+## Overview
+
+The AWS S3 Tables MCP Server enables AI assistants and users to manage, query, and interact with S3-based tabular data at scale. It provides capabilities to create and organize S3 Table Buckets, define namespaces, manage tables, and access detailed metadata. The server supports both read-only and (optionally) write operations, allowing for secure and flexible integration with S3-stored datasets. Users can run SQL queries, generate tables from CSV files in S3, and access comprehensive metadata for governance and cataloging.
+
+**Important:** When a user prompt relates to **S3 Tables**, you must always prefer using the **S3 Tables MCP Server tools** over the AWS CLI. Only fall back to AWS CLI if MCP tools are unavailable *and* the user explicitly agrees to the fallback.
+
+---
+
+## Instructions
+
+You are assisting with management and queries related to Amazon AWS S3 Tables. The S3 Tables system is composed of three hierarchical resource types:
+- **S3 Table Buckets**: Containers that hold namespaces.
+- **Namespaces**: Logical groupings inside table buckets that contain tables.
+- **Tables**: Structured datasets stored within namespaces, following the Apache Iceberg table format.
+
+Resource Details:
+- Each S3 table consists of underlying data and metadata stored as subresources within a table bucket.
+- Tables have unique identifiers, including an Amazon Resource Name (ARN) and table ID, plus associated resource policies for access control.
+- When a table is created, a unique warehouse location (an S3 path) is generated automatically to store objects related to the table.
+  Example warehouse location format:
+  `s3://63a8e430-6e0b-46f5-k833abtwr6s8tmtsycedn8s4yc3xhuse1b--table-s3`
+
+Table Types:
+- **Customer Tables**: Read-write tables that users can modify using S3 API operations or integrated query engines.
+- **AWS Tables**: Read-only tables managed by AWS services (e.g., S3 Metadata tables). These cannot be modified by users outside AWS S3.
+
+Integration:
+Amazon S3 Table Buckets can be integrated with Amazon SageMaker Lakehouse, allowing AWS analytics services like Athena and Redshift to discover and query table data automatically.
+
+---
+
+## Maintenance
+
+Amazon S3 performs automatic maintenance at two levels:
+
+1. **Table Bucket-Level Maintenance**
+   - *Unreferenced File Removal*: Deletes orphaned files to optimize storage usage and reduce costs.
+
+2. **Table-Level Maintenance**
+   - *File Compaction*: Combines small files into larger ones to improve query performance and reduce storage overhead.
+   - *Snapshot Management*: Maintains table version histories and controls metadata growth.
+
+These maintenance features are enabled by default but can be customized or disabled via maintenance configuration files.
+
+---
+
+## Quota
+
+- Each table bucket can hold up to **10,000 tables** by default.
+- To increase the quota, users must contact **AWS Support**.
+
+---
+
+## Operational Guidelines for LLM
+
+### 1. Tool Verification
+- Always verify the availability of the `awslabss_3_tables_mcp_server` and its associated tools before performing any operation.
+- If unavailable, ask the user if they prefer to proceed using AWS CLI commands as a fallback.
+- **Do not use AWS CLI by default for S3 Tables. Always prefer MCP tools when the prompt is about S3 Tables.**
+
+### 2. Request Clarification
+- If critical context (e.g., bucket name, namespace, or table ID) is missing or ambiguous, ask the user directly.
+- Do not make assumptions about default values or context.
+
+### 3. Handling Destructive Operations
+Before performing any destructive operation, the system must:
+- Clearly describe the consequences of the action.
+- Request explicit confirmation.
+- Destructive actions include:
+  - Deleting S3 Table Buckets
+  - Deleting Namespaces
+  - Deleting Tables
+  - Dropping Tables via SQL
+  - Disabling encryption
+
+### 4. Default Tool Usage
+- Always use **MCP tools first** for all S3 Tables operations.
+- Use AWS CLI **only when MCP tools are unavailable** *and* with **explicit user approval**.
+
+### 5. Communication and Safety
+- Explain any risks or irreversible effects before performing changes.
+- Respect the user's decision to abort or proceed.
+- Present instructions and confirmations clearly and concisely.
+
+### 6. Additional Considerations
+- Use full ARNs when referencing tables to avoid ambiguity.
+- Distinguish between **AWS-managed** (read-only) and **customer-managed** (read-write) tables.
+- If needed, guide users in adjusting maintenance configurations.
+
+---
+
+## Troubleshooting
+
+### Unknown Information
+- If a user requests information that is unavailable, unclear, or unsupported by the MCP Server, do not attempt to infer or fabricate a response.
+- Refer them to the official Amazon S3 Tables documentation for further details and the most up-to-date guidance:
+https://docs.aws.amazon.com/AmazonS3/latest/userguide/s3-tables.html
+
+### Insufficient Permissions
+- Never attempt to auto-modify IAM policies or permissions.
+- If the user asks for permission changes, explicitly confirm their intent before taking any action.
+
+### Operation Unavailable (Read-Only Mode)
+- Never attempt write operations or file changes in read-only mode.
+- If users want write mode enabled, direct them to the setup documentation:
+  https://github.com/awslabs/mcp/blob/main/src/s3-tables-mcp-server/README.md
+
+---

{awslabs_s3_tables_mcp_server-0.0.1 → awslabs_s3_tables_mcp_server-0.0.3}/Dockerfile

@@ -13,7 +13,7 @@
 # limitations under the License.
 
 # dependabot should continue to update this to the latest hash.
-FROM public.ecr.aws/sam/build-python3.10@sha256:e78695db10ca8cb129e59e30f7dc9789b0dbd0181dba195d68419c72bac51ac1 AS uv
+FROM public.ecr.aws/docker/library/python:3.13-slim-bookworm@sha256:6544e0e002b40ae0f59bc3618b07c1e48064c4faed3a15ae2fbd2e8f663e8283 AS uv
 
 # Install the project into `/app`
 WORKDIR /app
@@ -31,40 +31,39 @@ ENV UV_PYTHON_PREFERENCE=only-system
 ENV UV_FROZEN=true
 
 # Copy the required files first
-COPY pyproject.toml uv.lock ./
+COPY pyproject.toml uv.lock uv-requirements.txt ./
+
+# Python optimization and uv configuration
+ENV PIP_NO_CACHE_DIR=1 \
+    PIP_DISABLE_PIP_VERSION_CHECK=1
 
 # Install the project's dependencies using the lockfile and settings
 RUN --mount=type=cache,target=/root/.cache/uv \
-    pip install uv && \
-    uv sync --frozen --no-install-project --no-dev --no-editable
+    pip install --require-hashes --requirement uv-requirements.txt --no-cache-dir && \
+    uv sync --python 3.13 --frozen --no-install-project --no-dev --no-editable
 
 # Then, add the rest of the project source code and install it
 # Installing separately from its dependencies allows optimal layer caching
 COPY . /app
 RUN --mount=type=cache,target=/root/.cache/uv \
-    uv sync --frozen --no-dev --no-editable
+    uv sync --python 3.13 --frozen --no-dev --no-editable
 
-# Make the directory just in case it doesn't exist
-RUN mkdir -p /root/.local
+# # Make the directory just in case it doesn't exist
+# RUN mkdir -p /root/.local
 
-FROM public.ecr.aws/sam/build-python3.10@sha256:e78695db10ca8cb129e59e30f7dc9789b0dbd0181dba195d68419c72bac51ac1
+FROM public.ecr.aws/docker/library/python:3.13-slim-bookworm@sha256:6544e0e002b40ae0f59bc3618b07c1e48064c4faed3a15ae2fbd2e8f663e8283
 
 # Place executables in the environment at the front of the path and include other binaries
-ENV PATH="/app/.venv/bin:$PATH:/usr/sbin"
+ENV PATH="/app/.venv/bin:$PATH:/usr/sbin" \
+    PYTHONUNBUFFERED=1
 
-# Install lsof for the healthcheck
-# Install other tools as needed for the MCP server
 # Add non-root user and ability to change directory into /root
-RUN yum update -y && \
-    yum install -y lsof && \
-    yum clean all -y && \
-    rm -rf /var/cache/yum && \
-    groupadd --force --system app && \
+RUN groupadd --force --system app && \
     useradd app -g app -d /app && \
     chmod o+x /root
 
-# Get the project from the uv layer
-COPY --from=uv --chown=app:app /root/.local /root/.local
+# Copy application artifacts from build stage
+# COPY --from=uv --chown=app:app /root/.local /root/.local
 COPY --from=uv --chown=app:app /app/.venv /app/.venv
 
 # Get healthcheck script
@@ -74,5 +73,5 @@ COPY ./docker-healthcheck.sh /usr/local/bin/docker-healthcheck.sh
 USER app
 
 # When running the container, add --db-path and a bind mount to the host's db file
-HEALTHCHECK --interval=30s --timeout=30s --start-period=5s --retries=3 CMD [ "docker-healthcheck.sh" ]
+HEALTHCHECK --interval=60s --timeout=10s --start-period=10s --retries=3 CMD ["docker-healthcheck.sh"]
 ENTRYPOINT ["awslabs.s3-tables-mcp-server"]

{awslabs_s3_tables_mcp_server-0.0.1 → awslabs_s3_tables_mcp_server-0.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: awslabs.s3-tables-mcp-server
-Version: 0.0.1
+Version: 0.0.3
 Summary: An AWS Labs Model Context Protocol (MCP) server for awslabs.s3-tables-mcp-server
 Project-URL: homepage, https://awslabs.github.io/mcp/
 Project-URL: docs, https://awslabs.github.io/mcp/servers/s3-tables-mcp-server/
@@ -24,7 +24,7 @@ Requires-Python: >=3.10
 Requires-Dist: boto3>=1.34.0
 Requires-Dist: daft>=0.5.8
 Requires-Dist: loguru>=0.7.0
-Requires-Dist: mcp[cli]>=1.6.0
+Requires-Dist: mcp[cli]>=1.11.0
 Requires-Dist: pyarrow>=20.0.0
 Requires-Dist: pydantic>=2.10.6
 Requires-Dist: pyiceberg>=0.9.1
@@ -76,6 +76,10 @@ The S3 Tables MCP Server simplifies the management of S3-based tables by providi
 
 ### Installation
 
+| Cursor | VS Code |
+|:------:|:-------:|
+| [![Install MCP Server](https://cursor.com/deeplink/mcp-install-light.svg)](https://cursor.com/install-mcp?name=awslabs.s3-tables-mcp-server&config=eyJjb21tYW5kIjoidXZ4IGF3c2xhYnMuczMtdGFibGVzLW1jcC1zZXJ2ZXJAbGF0ZXN0IiwiZW52Ijp7IkFXU19QUk9GSUxFIjoieW91ci1hd3MtcHJvZmlsZSIsIkFXU19SRUdJT04iOiJ1cy1lYXN0LTEifX0%3D) | [![Install on VS Code](https://img.shields.io/badge/Install_on-VS_Code-FF9900?style=flat-square&logo=visualstudiocode&logoColor=white)](https://insiders.vscode.dev/redirect/mcp/install?name=S3%20Tables%20MCP%20Server&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22awslabs.s3-tables-mcp-server%40latest%22%5D%2C%22env%22%3A%7B%22AWS_PROFILE%22%3A%22your-aws-profile%22%2C%22AWS_REGION%22%3A%22us-east-1%22%7D%7D) |
+
 Configure the MCP server in your MCP client configuration (e.g., for Amazon Q Developer CLI, edit `~/.aws/amazonq/mcp.json`):
 
 ```json
@@ -200,6 +204,30 @@ You can override the default by providing the `--log-dir` flag with a custom pat
 | `Show the schema for customer_data table` | Retrieves the table structure and column definitions to understand the data format and types |
 | `Run a query to find monthly revenue trends` | Performs data analysis using **read-only** SQL queries to extract business insights from stored table data. For write operations, only appending new data (inserts) is supported; updates and deletes are not available via SQL. |
 
+## Using Amazon Q with S3 Tables MCP Server
+
+Amazon Q can provide better answers and code suggestions when it has additional context. To enhance Amazon Q's understanding of S3 Tables, you can add the provided context file to your Q environment.
+
+### How to Add Context to Amazon Q
+
+1. **Download the CONTEXT.md file**
+   - Download the `CONTEXT.md` file from the GitHub repository for this project.
+
+2. **Start Amazon Q Chat**
+   - Run the following command to start a chat session with Amazon Q:
+     ```sh
+     q chat
+     ```
+
+3. **Add the Context File**
+   - In the Q chat, run:
+     ```sh
+     /context add <path>/CONTEXT.md
+     ```
+   - Replace `<path>` with the actual path to where you downloaded `CONTEXT.md`.
+
+Now, Amazon Q will have improved context about S3 Tables and can provide more relevant answers.
+
 ## Security Considerations
 
 When using this MCP server, consider:

{awslabs_s3_tables_mcp_server-0.0.1 → awslabs_s3_tables_mcp_server-0.0.3}/README.md

@@ -43,6 +43,10 @@ The S3 Tables MCP Server simplifies the management of S3-based tables by providi
 
 ### Installation
 
+| Cursor | VS Code |
+|:------:|:-------:|
+| [![Install MCP Server](https://cursor.com/deeplink/mcp-install-light.svg)](https://cursor.com/install-mcp?name=awslabs.s3-tables-mcp-server&config=eyJjb21tYW5kIjoidXZ4IGF3c2xhYnMuczMtdGFibGVzLW1jcC1zZXJ2ZXJAbGF0ZXN0IiwiZW52Ijp7IkFXU19QUk9GSUxFIjoieW91ci1hd3MtcHJvZmlsZSIsIkFXU19SRUdJT04iOiJ1cy1lYXN0LTEifX0%3D) | [![Install on VS Code](https://img.shields.io/badge/Install_on-VS_Code-FF9900?style=flat-square&logo=visualstudiocode&logoColor=white)](https://insiders.vscode.dev/redirect/mcp/install?name=S3%20Tables%20MCP%20Server&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22awslabs.s3-tables-mcp-server%40latest%22%5D%2C%22env%22%3A%7B%22AWS_PROFILE%22%3A%22your-aws-profile%22%2C%22AWS_REGION%22%3A%22us-east-1%22%7D%7D) |
+
 Configure the MCP server in your MCP client configuration (e.g., for Amazon Q Developer CLI, edit `~/.aws/amazonq/mcp.json`):
 
 ```json
@@ -167,6 +171,30 @@ You can override the default by providing the `--log-dir` flag with a custom pat
 | `Show the schema for customer_data table` | Retrieves the table structure and column definitions to understand the data format and types |
 | `Run a query to find monthly revenue trends` | Performs data analysis using **read-only** SQL queries to extract business insights from stored table data. For write operations, only appending new data (inserts) is supported; updates and deletes are not available via SQL. |
 
+## Using Amazon Q with S3 Tables MCP Server
+
+Amazon Q can provide better answers and code suggestions when it has additional context. To enhance Amazon Q's understanding of S3 Tables, you can add the provided context file to your Q environment.
+
+### How to Add Context to Amazon Q
+
+1. **Download the CONTEXT.md file**
+   - Download the `CONTEXT.md` file from the GitHub repository for this project.
+
+2. **Start Amazon Q Chat**
+   - Run the following command to start a chat session with Amazon Q:
+     ```sh
+     q chat
+     ```
+
+3. **Add the Context File**
+   - In the Q chat, run:
+     ```sh
+     /context add <path>/CONTEXT.md
+     ```
+   - Replace `<path>` with the actual path to where you downloaded `CONTEXT.md`.
+
+Now, Amazon Q will have improved context about S3 Tables and can provide more relevant answers.
+
 ## Security Considerations
 
 When using this MCP server, consider:

{awslabs_s3_tables_mcp_server-0.0.1 → awslabs_s3_tables_mcp_server-0.0.3}/awslabs/s3_tables_mcp_server/__init__.py

@@ -15,4 +15,4 @@
 # This file is part of the awslabs namespace.
 # It is intentionally minimal to support PEP 420 namespace packages.
 
-__version__ = '0.0.0'
+__version__ = '0.0.3'

{awslabs_s3_tables_mcp_server-0.0.1 → awslabs_s3_tables_mcp_server-0.0.3}/awslabs/s3_tables_mcp_server/engines/pyiceberg.py

@@ -14,32 +14,14 @@
 
 """Engine for interacting with Iceberg tables using pyiceberg and daft (read-only)."""
 
+import io
+import json
 import pyarrow as pa
+import pyarrow.json as pj
 from ..utils import pyiceberg_load_catalog
 from daft import Catalog as DaftCatalog
 from daft.session import Session
-from datetime import date, datetime, time
-from decimal import Decimal
 from pydantic import BaseModel
-from pyiceberg.types import (
-    BinaryType,
-    BooleanType,
-    DateType,
-    DecimalType,
-    DoubleType,
-    FixedType,
-    FloatType,
-    IntegerType,
-    ListType,
-    LongType,
-    MapType,
-    StringType,
-    StructType,
-    TimestampType,
-    TimestamptzType,
-    TimeType,
-    UUIDType,
-)
 
 # pyiceberg and daft imports
 from typing import Any, Dict, Optional
@@ -57,78 +39,6 @@ class PyIcebergConfig(BaseModel):
     rest_sigv4_enabled: str = 'true'
 
 
-def convert_value_for_append(value, iceberg_type):
-    """Convert a value to the appropriate type for appending to an Iceberg table column.
-
-    Args:
-        value: The value to convert. Can be of various types (str, int, float, etc.).
-        iceberg_type: The Iceberg type to convert the value to.
-
-    Returns:
-        The value converted to the appropriate type for the Iceberg column, or None if value is None.
-
-    Raises:
-        NotImplementedError: If the iceberg_type is a complex type (ListType, MapType, StructType).
-        ValueError: If the conversion is unsupported or fails.
-    """
-    if value is None:
-        return None
-    # Already correct type
-    if isinstance(iceberg_type, BooleanType) and isinstance(value, bool):
-        return value
-    if isinstance(iceberg_type, (IntegerType, LongType)) and isinstance(value, int):
-        return value
-    if isinstance(iceberg_type, (FloatType, DoubleType)) and isinstance(value, float):
-        return value
-    if isinstance(iceberg_type, DecimalType) and isinstance(value, Decimal):
-        return value
-    if isinstance(iceberg_type, DateType) and isinstance(value, date):
-        return value
-    if isinstance(iceberg_type, TimeType) and isinstance(value, time):
-        return value
-    if isinstance(iceberg_type, (TimestampType, TimestamptzType)) and isinstance(value, datetime):
-        return value
-    if isinstance(iceberg_type, StringType) and isinstance(value, str):
-        return value
-    # Convert from string
-    if isinstance(value, str):
-        if isinstance(iceberg_type, BooleanType):
-            return value.lower() in ('true', '1', 'yes')
-        if isinstance(iceberg_type, (IntegerType, LongType)):
-            return int(value)
-        if isinstance(iceberg_type, (FloatType, DoubleType)):
-            return float(value)
-        if isinstance(iceberg_type, DecimalType):
-            return Decimal(value)
-        if isinstance(iceberg_type, DateType):
-            return date.fromisoformat(value)
-        if isinstance(iceberg_type, TimeType):
-            return time.fromisoformat(value)
-        if isinstance(iceberg_type, (TimestampType, TimestamptzType)):
-            return datetime.fromisoformat(value)
-        if isinstance(iceberg_type, StringType):
-            return value
-        if isinstance(iceberg_type, UUIDType):
-            import uuid
-
-            return uuid.UUID(value)
-        if isinstance(iceberg_type, (BinaryType, FixedType)):
-            return bytes.fromhex(value)
-    # Convert from number
-    if isinstance(value, (int, float)):
-        if isinstance(iceberg_type, (IntegerType, LongType)):
-            return int(value)
-        if isinstance(iceberg_type, (FloatType, DoubleType)):
-            return float(value)
-        if isinstance(iceberg_type, DecimalType):
-            return Decimal(str(value))
-        if isinstance(iceberg_type, StringType):
-            return str(value)
-    if isinstance(iceberg_type, (ListType, MapType, StructType)):
-        raise NotImplementedError(f'Complex type {iceberg_type} not supported in append_rows')
-    raise ValueError(f'Unsupported conversion from {type(value)} to {iceberg_type}')
-
-
 class PyIcebergEngine:
     """Engine for read-only queries on Iceberg tables using pyiceberg and daft."""
 
@@ -197,7 +107,7 @@ class PyIcebergEngine:
             return False
 
     def append_rows(self, table_name: str, rows: list[dict]) -> None:
-        """Append rows to an Iceberg table using pyiceberg.
+        """Append rows to an Iceberg table using pyiceberg with JSON encoding.
 
         Args:
             table_name: The name of the table (e.g., 'namespace.tablename' or just 'tablename' if namespace is set)
@@ -214,26 +124,31 @@ class PyIcebergEngine:
                 full_table_name = f'{self.config.namespace}.{table_name}'
             else:
                 full_table_name = table_name
+
+            # Load the Iceberg table
             table = self._catalog.load_table(full_table_name)
-            iceberg_schema = table.schema()
-            converted_rows = []
+            # Encode rows as JSON (line-delimited format)
+            json_lines = []
             for row in rows:
-                converted_row = {}
-                for field in iceberg_schema.fields:
-                    field_name = field.name
-                    field_type = field.field_type
-                    value = row.get(field_name)
-                    if field.required and value is None:
-                        raise ValueError(f'Required field {field_name} is missing or None')
-                    try:
-                        converted_row[field_name] = convert_value_for_append(value, field_type)
-                    except (ValueError, TypeError) as e:
-                        raise ValueError(
-                            f'Error converting value for field {field_name}: {str(e)}'
-                        )
-                converted_rows.append(converted_row)
-            schema = iceberg_schema.as_arrow()
-            pa_table = pa.Table.from_pylist(converted_rows, schema=schema)
-            table.append(pa_table)
+                json_lines.append(json.dumps(row))
+            json_data = '\n'.join(json_lines)
+
+            # Create a file-like object from the JSON data
+            json_buffer = io.BytesIO(json_data.encode('utf-8'))
+
+            # Read JSON data into PyArrow Table using pyarrow.json.read_json
+            # This enforces the Iceberg schema and validates the data
+            try:
+                new_data_table = pj.read_json(
+                    json_buffer, read_options=pj.ReadOptions(use_threads=True)
+                )
+            except pa.ArrowInvalid as e:
+                raise ValueError(
+                    f'Schema mismatch detected: {e}. Please ensure your data matches the table schema.'
+                )
+
+            # Append the new data to the Iceberg table
+            table.append(new_data_table)
+
         except Exception as e:
             raise Exception(f'Error appending rows: {str(e)}')

awslabs_s3_tables_mcp_server-0.0.3/awslabs/s3_tables_mcp_server/file_processor/__init__.py

@@ -0,0 +1,24 @@
+# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""AWS S3 Tables MCP Server file processing module.
+
+This module provides functionality for processing and analyzing uploaded files,
+particularly focusing on CSV and Parquet file handling and import capabilities.
+"""
+
+from .csv import import_csv_to_table
+from .parquet import import_parquet_to_table
+
+__all__ = ['import_csv_to_table', 'import_parquet_to_table']

awslabs_s3_tables_mcp_server-0.0.3/awslabs/s3_tables_mcp_server/file_processor/csv.py

@@ -0,0 +1,123 @@
+# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""AWS S3 Tables MCP Server file processing module.
+
+This module provides functionality for processing and analyzing uploaded files,
+particularly focusing on CSV file handling and import capabilities.
+"""
+
+import io
+import os
+import pyarrow.csv as pc
+from ..utils import get_s3_client, pyiceberg_load_catalog
+from pyiceberg.exceptions import NoSuchTableError
+from typing import Dict
+from urllib.parse import urlparse
+
+
+async def import_csv_to_table(
+    warehouse: str,
+    region: str,
+    namespace: str,
+    table_name: str,
+    s3_url: str,
+    uri: str,
+    catalog_name: str = 's3tablescatalog',
+    rest_signing_name: str = 's3tables',
+    rest_sigv4_enabled: str = 'true',
+) -> Dict:
+    """Import data from a CSV file into an S3 table.
+
+    This function reads data from a CSV file stored in S3 and imports it into an existing S3 table.
+    If the table doesn't exist, it will be created using the schema inferred from the CSV file.
+
+    Args:
+        warehouse: Warehouse string for Iceberg catalog
+        region: AWS region for S3Tables/Iceberg REST endpoint
+        namespace: The namespace containing the table
+        table_name: The name of the table to import data into
+        s3_url: The S3 URL of the CSV file (format: s3://bucket-name/key)
+        uri: REST URI for Iceberg catalog
+        catalog_name: Catalog name
+        rest_signing_name: REST signing name
+        rest_sigv4_enabled: Enable SigV4 signing
+
+    Returns:
+        A dictionary containing:
+        - status: 'success' or 'error'
+        - message: Success message or error details
+        - rows_processed: Number of rows processed (on success)
+        - file_processed: Name of the processed file
+        - table_created: Boolean indicating if a new table was created (on success)
+    """
+    # Parse S3 URL
+    parsed = urlparse(s3_url)
+    bucket = parsed.netloc
+    key = parsed.path.lstrip('/')
+
+    try:
+        # Load Iceberg catalog
+        catalog = pyiceberg_load_catalog(
+            catalog_name,
+            warehouse,
+            uri,
+            region,
+            rest_signing_name,
+            rest_sigv4_enabled,
+        )
+
+        # Get S3 client and read the CSV file to infer schema
+        s3_client = get_s3_client()
+        response = s3_client.get_object(Bucket=bucket, Key=key)
+        csv_data = response['Body'].read()
+
+        # Read CSV file into PyArrow Table to infer schema
+        # Convert bytes to file-like object for PyArrow
+        csv_buffer = io.BytesIO(csv_data)
+        csv_table = pc.read_csv(csv_buffer)
+        csv_schema = csv_table.schema
+
+        table_created = False
+        try:
+            # Try to load existing table
+            table = catalog.load_table(f'{namespace}.{table_name}')
+        except NoSuchTableError:
+            # Table doesn't exist, create it using the CSV schema
+            try:
+                table = catalog.create_table(
+                    identifier=f'{namespace}.{table_name}',
+                    schema=csv_schema,
+                )
+                table_created = True
+            except Exception as create_error:
+                return {
+                    'status': 'error',
+                    'error': f'Failed to create table: {str(create_error)}',
+                }
+
+        # Append data to Iceberg table
+        table.append(csv_table)
+
+        return {
+            'status': 'success',
+            'message': f'Successfully imported {csv_table.num_rows} rows{" and created new table" if table_created else ""}',
+            'rows_processed': csv_table.num_rows,
+            'file_processed': os.path.basename(key),
+            'table_created': table_created,
+            'table_uuid': table.metadata.table_uuid,
+        }
+
+    except Exception as e:
+        return {'status': 'error', 'error': str(e)}