# Changelog of Python Code Runner - pip, pandas & Dataset In/Out (`automa-flow/python-code-runner`) Actor

- **URL**: https://apify.com/automa-flow/python-code-runner/changelog.md
- **Full Actor documentation**: https://apify.com/automa-flow/python-code-runner.md

## Changelog

### 2026-10-18: First version (0.1)

#### Features

- Run your own Python 3.12 `main()` on rows in one call: inline `rows` (for example the items of an n8n or Make step), another Actor's Dataset (`datasetId`, `datasetOffset`, `maxRows`, `fields`) or only `input` parameters as a script.
- Two call modes: `allItems` calls `main(rows, ctx)` once; `eachItem` calls `main(row, ctx)` per row, so one failing row never stops the others.
- pandas, numpy, openpyxl and 13 more packages preinstalled; extra PyPI wheels from a `requirements` field, checked line by line before anything is installed.
- Encrypted `secrets` reach the code as environment variables and are redacted from logs, status rows and `RUN_SUMMARY`; the code text is never logged.
- Returned rows land in the Dataset in your own shape, with `_status` rows for failed or empty units, up to 10 saved files, and a `RUN_SUMMARY` with the input indexes to resend.
- Charges follow the work: one `code-run` per run whose code started, extra GB-minutes, one `packages-installed` and every started 100 rows beyond the first 100. Invalid input, failed installs and interrupted runs are not charged.