Skip to content

v4.2.0 - Optional model routing

Choose a tag to compare

View all tags
@vector233 vector233 released this 22 Jul 09:45
· 24 commits to main since this release
v4.2.0
1bda129
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Highlights

  • Adds opt-in Codex and Claude Code specialist model-routing overlays while keeping core-only installation as the default.
  • Preserves the user's root-session model; explorer and implementer use lower-cost models, while reviewer and architect use higher-capability models.
  • Renames the Companion Skill from ai-collaboration-workflow to repo-continuity.
  • Pins the Companion Skill bootstrap and Quick Start to v4.2.0.

Compatibility

  • Existing initialized repositories do not need to be initialized again.
  • Users upgrading the global Companion Skill should remove ai-collaboration-workflow, install repo-continuity, and invoke $repo-continuity.
  • Model routing remains disabled unless explicitly requested.
  • The Doctor and initialization cleanup still recognize the legacy payload marker for interrupted setups.
  • This release does not use GPT-5.3-Codex-Spark.

Validation

  • Full distribution lifecycle validation passes.
  • The official skills CLI installs repo-continuity from the public v4.2.0 tag.
  • The installed Skill bootstraps 22 core files from the pinned tag with no model-routing adapter by default.
  • Public opt-in Codex routing installs the expected .codex configuration and all four specialist agents.
Assets 2
Loading