Why

Every executable-source PR currently schedules all seven MATLAB compatibility shards. That is appropriate for ordinary development and compatibility-sensitive repairs, but it repeats unrelated platform sessions for a small hotfix whose repository-owned changed-file closure already names its complete evidence.

What changed

• Reserved hotfix/focused/<topic> as an explicit bounded-validation namespace.
• Added a clean Linux/latest MATLAB job that runs changedFast for eligible focused hotfixes.
• Kept documentation checks when focused hotfixes change authored docs.
• Made CI Gate require the focused job whenever selected.
• Kept ordinary hotfix/<topic> PRs on the full matrix.
• Made CI, Build, test-framework/catalog, maintainer-tool, resource, and dependency-governance paths automatically fall back to the full matrix even under the focused namespace.
• Added Python classifier, MATLAB repository architecture, policy documentation, and history evidence.

Evidence

python -m unittest discover -s .github/scripts -p 'test_*.py'
19 tests passed

labkittest.run(File=".github/workflows/ci.yml")
42 tests passed

labkittest.run(Owner="system/repository", Contract="system")
14 tests passed

matlab -batch "buildtool changedFast"
56 tests, 0 failed, 0 incomplete

matlab -batch "buildtool docsCheck"
376 generated files compared deterministically

Ruby YAML parse
ci.yml parsed successfully

check_integration_policy.py
Integration policy passed

Risks and follow-up

Focused validation is not a cross-platform compatibility claim. It must not be used when a repair depends on native dialogs, filesystems, graphics, packaging, external processes, or release-specific behavior. This governance PR deliberately receives the existing complete matrix before the focused route is used by a product hotfix.

Author confirmation

• The final diff matches the stated scope and contains no unrelated change.
• The evidence above covers required local and manual validation; every
  remaining unverified boundary is explicit.
• The diff contains only synthetic or generic public data, with no private
  App details, lab identifiers, local paths, or recognizable sample values.