terraform-multicloud
Copyright (c) 2026 SASVA AI Model Cognition Labs (MCL) Team

This DATASET — the compilation, the model-written `instruction` and `input`
fields, the provenance columns and ATTRIBUTION.csv — is licensed under the
Creative Commons Attribution 4.0 International Licence (CC-BY-4.0).

    https://creativecommons.org/licenses/by/4.0/legalcode

CC-BY-4.0 is inherited from the upstream source, which must be attributed:

    TerraDS: A Dataset for Terraform HCL Programs
    Zenodo, doi:10.5281/zenodo.20339474  (CC-BY-4.0)

------------------------------------------------------------------------------
THE CODE INSIDE REMAINS UNDER ITS ORIGINAL LICENCES
------------------------------------------------------------------------------

The `output` field of every row is a Terraform source file written by a third
party and redistributed here unaltered. It is NOT covered by the licence above.
Each row carries the `repo`, `license` and `file_path` of its origin, and
ATTRIBUTION.csv lists all 25,264 source repositories with their SPDX licence.

Row counts by source licence:

    MIT                 49,211      BSD-2-Clause          277
    Apache-2.0          24,750      WTFPL                 141
    UPL-1.0              1,222      ISC                    97
    Unlicense              937      0BSD                    8
    MIT-0                  904      BSD-3-Clause-Clear      1
    BSD-3-Clause           892
    CC0-1.0                472

76,450 rows (96.9%) are under licences requiring attribution and retention of
the licence text. If you redistribute this dataset or a derivative, carry
ATTRIBUTION.csv with it.

Apache-2.0 section 4(b) requires retention of NOTICE files. Repository-root
NOTICE files are NOT included here: the extraction unit is a single .tf file.
ATTRIBUTION.csv plus the `repo` column is the route back to them.

Removal requests: open a discussion on this dataset naming the repository.