Syntax
Python data pipelines,
without the boilerplate
Write data transformations in concise, readable Pivotal syntax and run against common Python data backends (Pandas, Polars, DuckDB).
with players
left merge results on player
filter minutes >= 900
goals_per_90 = goals / minutes * 90
with players as scoring_rates
group by player
agg mean goals_per_90
sort goals_per_90 desc
players = players.merge(results, how="left", on="player")
players = players[players["minutes"] >= 900].copy()
players["goals_per_90"] = players["goals"] / players["minutes"] * 90
scoring_rates = players.groupby("player")["goals_per_90"].mean()
scoring_rates = scoring_rates.reset_index()
scoring_rates = scoring_rates.sort_values("goals_per_90", ascending=False)
Install
Getting started
Install the Pivotal language and JupyterLab extension with pip, and the
VS Code extension
via marketplace.
pip install pivotal-lang pivotal-lab
MCP server
Generate Pivotal code with AI
Connect the public Pivotal MCP server to your AI agent, giving it access to concise Pivotal syntax descriptions, full Pivotal documentation, code generation checks and highlighting. Local MCP servers expose further execution and comparison tools for private data.
Watch the Claude MCP demo