Installation

Requirements

• Python 3.10 or newer (3.11 and 3.12 also supported)
• pip, uv, or any PEP 517-compatible installer

Install from PyPI

pip install dataspoc-pipe

Or with uv:

uv pip install dataspoc-pipe

Cloud storage extras

By default, Pipe writes to the local filesystem. To write to cloud buckets, install the appropriate extra:

Extra	Provider	Backend package
s3	Amazon S3	s3fs
gcs	Google Cloud Storage	gcsfs
azure	Azure Blob Storage	adlfs

# Amazon S3
pip install dataspoc-pipe[s3]

# Google Cloud Storage
pip install dataspoc-pipe[gcs]

# Azure Blob Storage
pip install dataspoc-pipe[azure]

# Multiple extras
pip install dataspoc-pipe[s3,gcs]

MCP server extra

To use Pipe as an MCP server for AI agent integration:

pip install dataspoc-pipe[mcp]

Installing Singer taps

Singer taps are separate packages. Install the ones you need:

# CSV files
pip install tap-csv

# PostgreSQL
pip install tap-postgres

# MySQL
pip install tap-mysql

# GitHub API
pip install tap-github

# Google Sheets (public)
# Built-in — no install needed

Each tap is an independent CLI tool. Pipe runs them as subprocesses, so they just need to be available in your PATH.

Tip

Use a virtual environment to keep tap dependencies isolated:

python -m venv .venv
source .venv/bin/activate
pip install dataspoc-pipe[s3] tap-csv tap-postgres

Verify the installation

dataspoc-pipe --version

Expected output:

dataspoc-pipe 0.2.0

Next steps

Run dataspoc-pipe init to create the configuration directory, then head to the Quickstart to build your first pipeline.