Skip to main content

Oxidizer Lite

Oxidizer Lite helps you turn Python functions into data workflows. You describe what runs, in what order, and where the data comes from and goes in YAML. The controller keeps things moving, while workers process batches and save checkpoints along the way. Use it to pull data from APIs or object storage, pass records through Redis or Kafka, and write results to Iceberg tables.

Active development: interfaces and configuration may change between releases.

Key Concepts

Concept Purpose
Controller Starts runs, schedules ready tasks, and tracks progress and checkpoints.
Worker Executes Python task functions and handles configured inputs and outputs.
API / MCP (Topology) Exposes run inspection, diagnostics, table queries, and configuration management to HTTP and MCP clients. The REST API also starts workflows.
Workflow (formula) YAML describing task nodes, dependencies, and input/output methods.
Connection Named backend settings referenced by workflows, separate from task logic.
Schema Defines table columns, record identity, and write policies such as upsert or SCD2 history.

See workflows and connections, schema configuration, and run state and checkpoints.

Quickstart

The supplied Redis demo reads a million-row Parquet fixture and writes Iceberg upsert and history tables using local RustFS storage.

From the repository root:

cd examples/quickstart
docker compose -f compose.redis.yml up -d
docker compose -f compose.redis.yml logs seed

Wait for Ready to invoke quickstart_observations_redis in the seed log. The first API image build downloads its documentation-search model; serving does not require a model download or API key.

REST API

Check readiness, then invoke the workflow:

curl --fail http://localhost:8027/api/health
curl --fail http://localhost:8027/api/docs/status

curl --fail-with-body --silent --show-error \
  --request POST http://localhost:8027/api/invocations \
  --header 'Content-Type: application/json' \
  --data '{"formula_id":"quickstart_observations_redis"}'

Use the returned run_id to inspect progress:

curl --fail-with-body --silent --show-error \
  "http://localhost:8027/api/formulas/quickstart_observations_redis/runs/<run-id>"

The quickstart guide includes result queries, expected counts, the Kafka alternative, customization, and cleanup instructions.

MCP

Connect an MCP client using Streamable HTTP at http://localhost:8027/mcp. Example tool calls (not REST requests):

{"name":"list_formulas","arguments":{}}
{"name":"get_formula_state","arguments":{"formula_id":"quickstart_observations_redis","run_id":"<run-id>"}}
{"name":"search_docs","arguments":{"query":"How do I troubleshoot a stalled workflow?","limit":3}}

See the API and MCP reference for routes, tools, and documentation lookup.

The API/MCP server has no built-in authentication. Keep it on loopback or behind trusted access control. Quickstart credentials and storage are disposable demo settings, not production configuration.

Documentation

Metadata

Release files for oxidizer-lite 0.2.3

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for oxidizer-lite 0.2.3
File Size Uploaded Provenance Actions
oxidizer_lite-0.2.3.tar.gz 141.0 kB With provenance Details for oxidizer_lite-0.2.3.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for oxidizer-lite 0.2.3
File Interpreter ABI Platform Actions
oxidizer_lite-0.2.3-py3-none-any.whl Python 3 none any Details for oxidizer_lite-0.2.3-py3-none-any.whl

Total release size: 302.2 kB

Release files / oxidizer_lite-0.2.3.tar.gz

Download URL oxidizer_lite-0.2.3.tar.gz
Size 141.0 kB
Tags Source
SHA-256 checksum
How to use checksums
3a8781f220d9833a3dab8958d7c63f1e78a2b4c1533b7c8b40131e5e958da33f
BLAKE2b-256 checksum
How to use checksums
1e2861c2053161417890fb9e52e9be19d1a80605d39455fbf06ef2c774e7c745
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 2, 2026.

Transparency log

Release files / oxidizer_lite-0.2.3-py3-none-any.whl

Download URL oxidizer_lite-0.2.3-py3-none-any.whl
Size 161.3 kB
Tags Python 3
SHA-256 checksum
How to use checksums
9ad884627bbd219f889bdd73c9039a3beff7f3065e7219af2167bd11602e1fd3
BLAKE2b-256 checksum
How to use checksums
c249ba7f4cdafc1e638d32507d33d60f295a45ee5b5d7a4b6c4ac03460caaa0a
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 2, 2026.

Transparency log

Security for oxidizer-lite 0.2.3

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

All 2 files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for oxidizer_lite-0.2.3.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 2, 2026.

Transparency log

A 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.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page