mostlyai-mock 0.0.10__py3-none-any.whl → 0.0.12__py3-none-any.whl

mostlyai/mock/__init__.py

@@ -15,4 +15,4 @@
 from mostlyai.mock.core import sample
 
 __all__ = ["sample"]
-__version__ = "0.0.10"  # Do not set this manually. Use poetry version [params].
+__version__ = "0.0.12"  # Do not set this manually. Use poetry version [params].

mostlyai/mock/core.py

@@ -18,6 +18,7 @@ import json
 from collections import deque
 from collections.abc import Generator
 from enum import Enum
+import os
 from typing import Any, Literal, Type
 
 import litellm
@@ -265,7 +266,7 @@ def _create_table_prompt(
         prompt += f"## Context Table Primary Key: `{primary_keys[fk.referenced_table]}`\n\n"
 
         prompt += f"## Context Table Data:\n\n"
-        prompt += f"{context_data.to_json(orient='records', indent=2)}\n\n"
+        prompt += f"{context_data.to_json(orient='records', date_format='iso', indent=2)}\n\n"
 
     # add non-context table names, primary keys and data
     if foreign_keys and len(foreign_keys) > 1:
@@ -279,7 +280,7 @@ def _create_table_prompt(
             prompt += f"## Non-Context Table Primary Key: `{primary_keys[fk.referenced_table]}`\n\n"
 
             prompt += f"## Non-Context Table Data:\n\n"
-            prompt += f"{non_context_data[fk.referenced_table].to_json(orient='records', indent=2)}\n\n"
+            prompt += f"{non_context_data[fk.referenced_table].to_json(orient='records', date_format='iso', indent=2)}\n\n"
 
     # add instructions
     prompt += "\n## Instructions:\n\n"
@@ -384,12 +385,12 @@ def _create_table_rows_generator(
                 for i in range(0, len(data), batch_size):
                     yield data.iloc[i : i + batch_size]
 
-    # ensure model supports response_format and json schema
-    supported_params = litellm.get_supported_openai_params(model=llm_config.model)
-    assert "response_format" in supported_params
-    assert litellm.supports_response_schema(llm_config.model), (
-        "The model does not support structured output / JSON mode."
-    )
+    if not llm_config.model.startswith("litellm_proxy/"):
+        # ensure model supports response_format and json schema (this check does not work with litellm_proxy)
+        supported_params = (litellm.get_supported_openai_params(model=llm_config.model) or [])
+        assert "response_format" in supported_params and litellm.supports_response_schema(llm_config.model), (
+            "The model does not support structured output / JSON mode."
+        )
 
     # derive context data (if first foreign key is present) and harmonize sample size accordingly
     context_data: pd.DataFrame | None = None
@@ -398,6 +399,7 @@ def _create_table_rows_generator(
         assert generated_data is not None
         assert context_table_name in generated_data
         context_data = generated_data[context_table_name]
+        batch_size = 1  # generate one sequence at a time
         sample_size = len(context_data)
 
     # derive non-context data (if more than one foreign key is present)

mostlyai/mock/mcp_server.py

@@ -1,19 +1,21 @@
 import os
 import tempfile
-import zipfile
 
+import pandas as pd
 from fastmcp import Context, FastMCP
 
 from mostlyai import mock
 
 SAMPLE_MOCK_TOOL_DESCRIPTION = f"""
-It is proxy to the `mostlyai.mock.sample` function.
+This tool is a proxy to the `mostlyai.mock.sample` function.
+It returns a dictionary. The keys are the table names, the values are the Paths to the generated CSV files.
 
-This tool returns an URL or a Path to the generated CSV bundle (as ZIP file).
-Present the result nicely to the user, in Markdown format. Some examples:
+Present the result nicely to the user, in Markdown format. Example:
+
+Mock data can be found under the following paths:
+- `/tmp/tmpl41bwa6n/players.csv`
+- `/tmp/tmpl41bwa6n/seasons.csv`
 
-"Mock data is ready to download: [Mock Data](https://example.com/mock_data.zip)" (if result is a link)
-"Mock data can be found in `/tmp/tmpl41bwa6n/mock_data.zip`" (if result is a path)
 
 What comes after the `=============================` is the documentation of the `mostlyai.mock.sample` function.
 
@@ -24,20 +26,18 @@ What comes after the `=============================` is the documentation of the
 mcp = FastMCP(name="MostlyAI Mock MCP Server")
 
 
-def _store_locally(data: dict) -> str:
+def _store_locally(data: dict[str, pd.DataFrame]) -> dict[str, str]:
     temp_dir = tempfile.mkdtemp()
-    zip_path = os.path.join(temp_dir, "mock_data.zip")
-    with zipfile.ZipFile(zip_path, "w") as zip_file:
-        for table_name, df in data.items():
-            csv_path = os.path.join(temp_dir, f"{table_name}.csv")
-            df.to_csv(csv_path, index=False)
-            zip_file.write(csv_path, arcname=f"{table_name}.csv")
-
-    return os.path.abspath(zip_path)
+    locations = {}
+    for table_name, df in data.items():
+        csv_path = os.path.join(temp_dir, f"{table_name}.csv")
+        df.to_csv(csv_path, index=False)
+        locations[table_name] = csv_path
+    return locations
 
 
 @mcp.tool(description=SAMPLE_MOCK_TOOL_DESCRIPTION)
-def sample_mock_data(
+def mock_data(
     *,
     tables: dict[str, dict],
     sample_size: int,
@@ -46,7 +46,7 @@ def sample_mock_data(
     temperature: float = 1.0,
     top_p: float = 0.95,
     ctx: Context,
-) -> str:
+) -> dict[str, str]:
     # Notes:
     # 1. Returning DataFrames directly results in converting them into truncated string.
     # 2. The logs / progress bars are not propagated to the MCP Client. There is a dedicated API to do that (e.g. `ctx.info(...)`)
@@ -63,8 +63,8 @@ def sample_mock_data(
         return_type="dict",
     )
     ctx.info(f"Generated mock data for `{len(tables)}` tables")
-    url = _store_locally(data)
-    return url
+    locations = _store_locally(data)
+    return locations
 
 
 def main():

{mostlyai_mock-0.0.10.dist-info → mostlyai_mock-0.0.12.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mostlyai-mock
-Version: 0.0.10
+Version: 0.0.12
 Summary: Synthetic Mock Data
 Project-URL: homepage, https://github.com/mostly-ai/mostlyai-mock
 Project-URL: repository, https://github.com/mostly-ai/mostlyai-mock
@@ -244,7 +244,7 @@ print(df)
 
 This repo comes with MCP Server. It can be easily consumed by any MCP Client by providing the following configuration:
 
-```yaml
+```json
 {
   "mcpServers": {
       "mostlyai-mock-mcp": {
@@ -266,5 +266,33 @@ For example:
 - in Cursor, go to "Settings" > "Cursor Settings" > "MCP" > "Add new global MCP server" and paste the above into `mcp.json`
 
 Troubleshooting:
-1. If MCP Server is not picked up by the MCP Client, specify full path in `command`, e.g. `/Users/johnsmith/.local/bin/uvx`
-2. MCP Inspector can be used for debugging: `npx @modelcontextprotocol/inspector -- uvx --from mostlyai-mock mcp-server`
+1. If the MCP Client fails to detect the MCP Server, provide the absolute path in the `command` field, for example: `/Users/johnsmith/.local/bin/uvx`
+2. To debug MCP Server issues, you can use MCP Inspector by running: `npx @modelcontextprotocol/inspector -- uvx --from mostlyai-mock mcp-server`
+3. In order to develop locally, modify the configuration by replacing `"command": "uv"` (or use the full path to `uv` if needed) and `"args": ["--directory", "/Users/johnsmith/mostlyai-mock", "run", "mcp-server"]`
+
+
+## LiteLLM Proxy Server
+
+In order to consume LiteLLM Proxy Server, the user must:
+- Set `LITELLM_PROXY_API_KEY` and `LITELLM_PROXY_API_BASE`
+- Prefix the `model` with LiteLLM Proxy Server provider: `litellm_proxy`. For example, `litellm_proxy/openai/gpt-4.1-nano`
+
+```python
+from mostlyai import mock
+import os
+
+os.environ["LITELLM_PROXY_API_BASE"] = "https://litellm-proxy-production-7a86.up.railway.app/"
+tables = {
+    "guests": {
+        "prompt": "Guests of an Alpine ski hotel in Austria",
+        "columns": {
+            "name": {"prompt": "first name and last name of the guest", "dtype": "string"},
+        },
+    }
+}
+df = mock.sample(tables=tables, sample_size=10, model="litellm_proxy/mostlyai/openai/gpt-4.1-nano")
+
+print(df)
+```
+
+Read more [here](https://docs.litellm.ai/docs/providers/litellm_proxy).

mostlyai_mock-0.0.12.dist-info/RECORD

@@ -0,0 +1,8 @@
+mostlyai/mock/__init__.py,sha256=EzC1pduN2tfMeNk5Q4DHpNCZ8Erlo4KdsEoiZSq76GY,715
+mostlyai/mock/core.py,sha256=OpVipCV-7t9WtdQ7NDs2Nl58Y_9jsLhwqT8vFrQKgLM,30115
+mostlyai/mock/mcp_server.py,sha256=FqtgGdtuncpLcPySQk2V2RXASFvfV27rqlrnXnmSE7M,2311
+mostlyai_mock-0.0.12.dist-info/METADATA,sha256=O7ACxfwzQE7M8pipE1kLwf-sdWkIaAZzBfHFOS1vwA8,13526
+mostlyai_mock-0.0.12.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+mostlyai_mock-0.0.12.dist-info/entry_points.txt,sha256=XDbppUIAaCWW0nresVep8zb71pkzZuFA16jCBHq8CU8,61
+mostlyai_mock-0.0.12.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+mostlyai_mock-0.0.12.dist-info/RECORD,,

mostlyai_mock-0.0.10.dist-info/RECORD

@@ -1,8 +0,0 @@
-mostlyai/mock/__init__.py,sha256=grEzN1CxhOdKf-ow8fWVNTzUrB2Ogg8IBQIKqyaOc7I,715
-mostlyai/mock/core.py,sha256=p5VAsRppzAc4P8FqKEunfQ3cPjImUU2cEc6yqHJVhMg,29884
-mostlyai/mock/mcp_server.py,sha256=juy5n6-Xo-ZVC3u5o2zylLgN1CaAwzIaS_bk2qXxpcU,2439
-mostlyai_mock-0.0.10.dist-info/METADATA,sha256=OHHgoJQ6_RL20Ze5ofFvMscXK5oBOXTfYyEYy0XwsCU,12443
-mostlyai_mock-0.0.10.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-mostlyai_mock-0.0.10.dist-info/entry_points.txt,sha256=XDbppUIAaCWW0nresVep8zb71pkzZuFA16jCBHq8CU8,61
-mostlyai_mock-0.0.10.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-mostlyai_mock-0.0.10.dist-info/RECORD,,