DataNyx
Let an AI agent work with your data without ever seeing it.
DataNyx turns files and databases into value-free schemas an agent can write code against, and replaces names and identifiers in free text with placeholders.
Getting started · Documentation · Command reference · Privacy model · Website · Changelog
You want an agent to write a report, a query or a data-cleaning script, but the data is customers, patients or payroll. Pasting rows into a chat is a leak, and describing the columns by hand is tedious. DataNyx sits in between:
| What the agent gets | What it never gets | |
|---|---|---|
Structured datadatanyx schema |
Field names, types, formats, nesting, primary and foreign keys | Values, examples, enum labels, counts or statistics |
Free textdatanyx anonymize |
The text with <PERSON_1>, <EMAIL_1> and friends in place of identifiers |
The identifiers themselves (best effort: review the result) |
A first look
$ datanyx schema customers.csv --format text
root: array<{
id: integer
email: string<email>
signed_up: string<date>
plan: string
lifetime_value: number
notes: string | null
}>
$ datanyx anonymize call-notes.txt
<PERSON_1> (<EMAIL_1>) called on <DATE_1> about invoice
<ID_1>. Ms. <PERSON_1> lives in <LOCATION_1>. Her manager <PERSON_2> agreed.
The agent sees the shape of the table and the story in the text, and none of the values. It can also try its code on fake rows generated from the schema.
Install
git clone <this repository> datanyx && cd datanyx
uv sync --all-extras # or: pip install -e ".[all]"
datanyx doctor # shows which optional packages and models are installed
Python 3.10 or newer. Supported sources lists the extra for each format.
What's in the documentation
| Getting started | Install, a first schema, a first anonymized text |
| Schemas, synthetic data | The core: what the agent sees instead of the data |
| Debugging, checking code | Fixing code without seeing the data |
| Scanning, anonymization, pseudonymization, detection quality | Finding and hiding personal data |
| Using it with an agent | datanyx protect, the guard hook and the MCP server |
| Installing and deploying | PyPI, Docker, a pre-commit hook, a GitHub Action, a shared MCP server |
| Team settings, schema diff | Shared policy, and change detection for CI |
| Privacy model | What is guaranteed, what isn't, and the known limits |
| Command reference, Python API | Every option, and use from code |
| Contributing | Architecture, adding a source, tests |
To read it as a site: uv sync --group docs then uv run mkdocs serve.
License
MIT © 2026 ludenberg
Metadata
Release files for datanyx 0.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| datanyx-0.2.0.tar.gz | 343.5 kB | With provenance | Details for datanyx-0.2.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| datanyx-0.2.0-py3-none-any.whl | Python 3 | none | any | Details for datanyx-0.2.0-py3-none-any.whl |
Total release size: 595.1 kB
Release files / datanyx-0.2.0.tar.gz
| Download URL | datanyx-0.2.0.tar.gz |
|---|---|
| Size | 343.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
376e73bdfd5e30877abc367b9852ae0a6856bb35b8265b31b981fc9feb527106
|
|
BLAKE2b-256 checksum How to use checksums |
11b186bf24c55ad1a47a9ca701a495a9265a36907921571bb9ad235ce4549918
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 6, 2026.
Transparency logRelease files / datanyx-0.2.0-py3-none-any.whl
| Download URL | datanyx-0.2.0-py3-none-any.whl |
|---|---|
| Size | 251.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
54ef6becba32e51909171d9baae037c5dad95ab17ab6ecdbc23bcc1fd75c2d45
|
|
BLAKE2b-256 checksum How to use checksums |
652ed11d629c2f466853fd5b75fa3a4321667ab0fc3dec1a0b5fae51acd18d32
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 6, 2026.
Transparency logSecurity for datanyx 0.2.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
Both files in this release have been uploaded with provenance metadata.
Attestation for datanyx-0.2.0.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 6, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.