Skip to content

Standardize public metadata, release qualification, docs, and sidecar #50

Closed

Description

@stefan-jansen

Objective

Bring ml4t-data, its public repository settings, release process, deployed documentation, and
private development sidecar into compliance with the current ML4T ecosystem standards.

Confirmed starting deviations

  • PyPI 0.1.3 publishes a generic team identity, incomplete keywords and project URLs, and a stale
    README description of the ecosystem.
  • Source metadata does not use the canonical author and maintainer identities, description,
    keywords, or project URLs.
  • The README lacks the complete link and development-gate contract and contains stale or
    unsupported claims.
  • The release repository does not contain the required one-line CLAUDE.md import and its root and
    nested agent instructions contain detailed, volatile repository and API inventories.
  • CI lacks part of the common pull-request gate and the release workflow does not yet implement the
    complete commit-bound artifact-manifest and documentation-publication sequence.
  • Deployed documentation lacks ml4t-library, ml4t-version, and ml4t-commit metadata.
  • Main-branch protection and repository security settings do not yet satisfy the complete shared
    intake and release contract.
  • The private ml4t-data-dev sidecar has no origin and has existing untracked content that must be
    classified and preserved while adding the required sidecar topology.

Acceptance criteria

  • Source, built wheel, GitHub settings, deployed documentation, and the next authorized PyPI
    release agree on the canonical metadata.
  • QLDM_DATA_ROOT and QldmError remain compatibility aliases with tested runtime deprecation
    warnings and documented replacements.
  • README installation and quick-start behavior pass from a clean wheel install and all links
    resolve.
  • Pull requests and releases run the required lint, format, type, test, build, security,
    documentation, and ecosystem qualification gates without weakening provider-specific checks.
  • Release artifacts are built once, bound to the candidate commit by a digest manifest, published
    through trusted publishing only after documentation deployment succeeds, and verified after
    publication.
  • The sidecar satisfies the private repository topology without losing existing work.
  • The full repository quality suite, strict documentation build, package checks, workflow/security
    validation, and ecosystem qualification pass.

Tracked by ml4t/ecosystem#13.

Activity

  1. added
    type: maintenanceInternal maintenance without user-visible behavior
    priority: highMaterial user impact requiring prompt remediation
    compatibility: affectedHas Python, platform, dependency, or cross-library compatibility impact
    ecosystemTracked by a shared ML4T ecosystem requirement
    on Sep 8, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    compatibility: affectedHas Python, platform, dependency, or cross-library compatibility impactecosystemTracked by a shared ML4T ecosystem requirementpriority: highMaterial user impact requiring prompt remediationstatus: acceptedAccepted for implementationtype: maintenanceInternal maintenance without user-visible behavior

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions