> ## Documentation Index
> Fetch the complete documentation index at: https://docs.datafog.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# 0.4.4

> Reduce phone and credit-card false positives in numeric data.

## Detection fixes

* Reject unprefixed North American phone candidates whose area or exchange code
  starts with 0 or 1. This suppresses common Unix timestamp false positives
  ([PR #41](https://github.com/DataFog/datafog-core/pull/41)).
* Reject credit-card candidates embedded in serialized decimal numbers while
  retaining formatted card detection after decimal-like prefixes
  ([PR #40](https://github.com/DataFog/datafog-core/pull/40)).

## Compatibility and limitations

These fixes apply to Rust, Python, Node.js, and browser WASM. Public APIs,
detector selection, and finding offset semantics are unchanged.

Phone detection remains permissive for candidates beginning with `+`; the
North American rule is not full international numbering-plan validation.
Numeric identifiers can still resemble real phone numbers. Metadata-based
validation remains tracked in
[issue #49](https://github.com/DataFog/datafog-core/issues/49).

The credit-card decimal guard cannot resolve every ambiguous numeric or CSV
context. Keep application-provided field context and detector selection explicit.

## Verification and publication

All runtime packages and lockfiles are aligned at **0.4.4**. Publication uses
`rust-v0.4.4`, `python-v0.4.4`, `node-v0.4.4`, and `wasm-v0.4.4` on the same
release commit. Candidate checks do not establish registry availability;
publication and clean registry installations must be verified separately.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.