Skip to content

chore: regenerate models from upstream schemas#124

Merged
stack72 merged 1 commit intomainfrom
automated/regenerate-models
May 5, 2026
Merged

chore: regenerate models from upstream schemas#124
stack72 merged 1 commit intomainfrom
automated/regenerate-models

Conversation

@stack72
Copy link
Copy Markdown
Contributor

@stack72 stack72 commented May 5, 2026

Summary

Automated regeneration of extension models from upstream provider schemas.

Changes

  • aws: 16 files changed
  • gcp: 13 files changed
  • digitalocean: 2 files changed

Schema Sources

  • AWS: CloudFormation Resource Schema
  • GCP: Google Discovery Documents
  • Hetzner: Hetzner Cloud OpenAPI spec
  • DigitalOcean: DigitalOcean OpenAPI spec

Review Notes

  • Files under model/ are auto-generated — review the manifest.yaml diffs for version changes
  • CalVer versioning with content-based change detection ensures versions only bump when content changes
  • Publishing happens automatically when this PR is merged (via the publish workflow)

Copy link
Copy Markdown
Contributor

@github-actions github-actions Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This is an automated regeneration of extension models from upstream provider schemas (AWS CloudFormation, GCP Discovery, DigitalOcean OpenAPI). All 31 changed files are under model/ with consistent auto-generated headers, CalVer version bumps, and proper upgrade entries.

Blocking Issues

None.

Suggestions

  1. model/aws/datazone/extensions/models/user_profile.ts — non-optional fields in nested StateSchema
    The two new fields added inside Details.Iam are not marked .optional():

    Iam: z.object({
      Arn: z.string(),
      SessionName: z.string(),    // NEW — required
      GroupProfileId: z.string(), // NEW — required
    }),

    If the AWS API doesn't always return SessionName/GroupProfileId (e.g. for non-session IAM profiles), Zod will reject the parse and sync/get will throw. Worth confirming the upstream CloudFormation schema marks these as always-present; if not, they should be .optional().

  2. model/aws/chime/README.md — now documents only app_instance_bot
    The README was regenerated to replace all app_instance references with app_instance_bot, but app_instance.ts still exists in the same package. The README is now misleading for users of the app_instance model. If the README generator is supposed to document a single model per run, this is a codegen-tooling gap worth tracking.

  3. GCP stale release notes removal — Several GCP manifests (accesscontextmanager, bigquerydatapolicy, bigtableadmin, cloudsearch, compute, networkservices, pubsub) have their previous releaseNotes blocks removed with no replacement. This appears to be intentional cleanup (each run only surfaces the current version's notes), but it means those manifests now carry no notes at all. Consistent behaviour across providers would be clearer.

Copy link
Copy Markdown
Contributor

@github-actions github-actions Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This is an automated regeneration of extension models from upstream provider schemas (AWS CloudFormation, GCP Discovery, DigitalOcean OpenAPI). All 31 changed files are under model/ with consistent auto-generated headers, CalVer version bumps, and proper upgrade entries.

Blocking Issues

None.

Suggestions

  1. model/aws/datazone/extensions/models/user_profile.ts — non-optional fields in nested StateSchema
    Two new fields added inside Details.Iam are not marked .optional(): SessionName and GroupProfileId. If the AWS API does not always return these (e.g. for non-session IAM profiles), Zod will reject the parse and sync/get will throw. Worth confirming the upstream CloudFormation schema marks these as always-present; if not, they should be .optional().

  2. model/aws/chime/README.md — now documents only app_instance_bot
    The README was regenerated replacing all app_instance references with app_instance_bot, but app_instance.ts still exists in the package. The README is now misleading for app_instance users. If the README generator documents a single model per run, this is a codegen tooling gap worth tracking.

  3. GCP stale release notes removal — Several GCP manifests (accesscontextmanager, bigquerydatapolicy, bigtableadmin, cloudsearch, compute, networkservices, pubsub) have their previous releaseNotes blocks removed with no replacement. This appears intentional (each run surfaces only the current version's notes), but it means those manifests now carry no notes at all. Consistent behaviour across providers would be clearer.

Copy link
Copy Markdown
Contributor

@github-actions github-actions Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review - No blocking issues

@stack72 stack72 merged commit f437401 into main May 5, 2026
32 checks passed
@stack72 stack72 deleted the automated/regenerate-models branch May 5, 2026 19:37
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant