dataframe-serialization

Rules for converting pandas DataFrames or Apache Arrow tables into records that APIs and frontends can safely read. The required helpers preserve dates as ISO-formatted text and handle unusual data types.

In plain words
What is it for?
Use it when returning table data from APIs, sending streaming events, or exposing DataFrame data to a frontend.
Why use it?
It prevents dates from becoming large timestamp numbers or other values that the frontend cannot display correctly.

Cursor rule for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/microsoft/data-formulator/dataframe-serialization
Clone the repo
git clone --depth 1 https://github.com/microsoft/data-formulator

Made for: Cursor.

Per session 369 This file is loaded in full into every session.
When invoked 369 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00369 $0.00369
Opus 5 $0.00185 $0.00185
Sonnet 5 $0.00074 $0.00074
Haiku 4.5 $0.00037 $0.00037

Measured yesterday against content hash 422675decc33, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dataframe-serialization scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursor/rules/dataframe-serialization.mdc · 49 lines

What it actually says

DataFrame Serialization

All DataFrame-to-records conversion for API responses, streaming events, or frontend-visible data MUST use the centralized helpers in data_formulator.datalake.parquet_utils:

Source type Helper
pd.DataFrame df_to_safe_records(df)
pa.Table (Arrow) get_sample_rows_from_arrow(table)

Why

pandas.DataFrame.to_json(orient='records') defaults to date_format='epoch', which serializes datetime columns as epoch milliseconds (e.g. 1773532800000). The frontend interprets these as plain numbers and renders them with commas (1,773,532,800,000) instead of formatted dates.

df_to_safe_records enforces date_format='iso' and default_handler=str, ensuring datetimes become ISO-8601 strings and exotic types degrade gracefully.

Banned Patterns

# BAD — missing date_format, datetimes become epoch numbers
json.loads(df.to_json(orient='records'))

# BAD — to_dict returns Timestamp objects, not JSON-safe values
df.to_dict(orient='records')

# ACCEPTABLE but should be unified for consistency
json.loads(df.to_json(orient='records', date_format='iso'))

Correct Pattern

from data_formulator.datalake.parquet_utils import df_to_safe_records

rows = df_to_safe_records(df)
preview = df_to_safe_records(df.head(5))

Exceptions

Internal data processing that never reaches the frontend or JSON serialization (e.g. Kusto SDK metadata parsing, Vega-Lite spec construction) may use to_dict(orient='records') directly.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 49 lines · 369 tokens per session scan A 422675decc33

Subscribe to this mod's changes

dataframe-serialization is a cursor rule published in the GitHub repository microsoft/data-formulator (17,048 stars, last pushed 3d ago), licensed MIT. It adds 369 tokens to every session, about $0.0018 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.