Metadata-Version: 2.4
Name: dsl-spa
Version: 0.3.14
Summary: A tool for building Structured Agentic Pipelines
Author-email: superwise-brian <brian.mcclannahan@superwise.ai>
License-Expression: MIT
Project-URL: Homepage, https://github.com/superwise-ai/dsl-spa
Project-URL: Issues, https://github.com/superwise-ai/dsl-spa/issues
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: OS Independent
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: setuptools>=77.0.3
Requires-Dist: pandas>=2.0.0
Requires-Dist: altair>=5.3.0
Requires-Dist: SQLAlchemy>=2.0.0
Requires-Dist: langchain-openai>=0.3.3
Requires-Dist: sentence-transformers>=3.4.1
Dynamic: license-file

# Domain Specific Language - Structured Pipeline Agent

DSL-SPA is an open-source Python library for connecting LLM Agents to ETL and other common tasks.  With this tool, you can connect LLM Agents to complex Data Pipelines.

Key Features of DSL-SPA

- Building SQL Queries - construct sql queries from fields extracted by Agents

- Applying Data Transformations - selectively apply data transformations based on Agent instructions

- Generating Unstructured Text Summaries - generate summaries of query results and data transformations 

- Generating Visualizations - generate vega-lite visualizations of query results and data transformations

- Executing Commands - generate command sequences or execute python fucntions

For some examples of what you can do with dsl-spa sdk check out the [streamlit demo](https://pipeline-agent.streamlit.app/).

## Getting Started

Get started with dsl-spa by installing the Python library via pip

```
pip install dsl-spa
```

A standard pipeline can be imported with

```
from dsl_spa.pipeline.pipeline import StandardPipeline
```

For more details on constructing a pipeline, reference [Building a Pipeline](https://github.com/superwise-ai/dsl-spa/blob/main/docs/Creating_a_Pipeline_Schema.md).

For more details on the flow of Semantic Caching using DSL-SPA, refer to [Semantic Cache Flow](https://github.com/superwise-ai/dsl-spa/blob/main/docs/semantic_caching_flow.jpg)
