Dependency-ready integration frontier

These examples and illustrative results are independently authored teaching materials, not measured model results.

Use case

Teaching S defines schema, E depends on S and C copy is independent. S may remain tracker-open until final PR closure even after verified integration. Readiness comes from consumable landed state, not closed labels.

Mechanism

Compute frontier from graph/integration, execute authorized independent work in isolation and serialize integration/checks before recomputing. S/C can run; E waits for integrated S. Shared files/interfaces may add dependencies/naming constraints. Missing ignored fixtures/credentials remain explicit gaps, not skipped-green claims.

Bad example

Start everything together or block E after S integration solely because tracker is open. Report green when missing local fixtures skip tests.

Good example

Run teaching S/C isolated from current integration state. After S lands and passes, start E from actual status rather than closure time. Integrate one result/check at a time and resolve shared fields. Missing required fixtures/credentials are unverified, with permitted main-environment checks where applicable, not guessed passes.

Why the change matters

Readiness needs actual implementation, not delayed tracker workflow. Isolation limits direct collision; serialized integration ties shared change to verified combinations.

Observable expectation

The fixture starts S/C, verifies integrated S, then E without premature work or tracker-induced indefinite blocking. Inspect matched revisions and separate skipped tests.

Isolation proves no merge-conflict absence; no real worktree/PR is created.

Limits

Isolation postpones shared collisions and textual edges may miss dependencies. Follow current restrictions on parallel/Git work; frozen reset/merge procedures grant no permission.