Pilot operate slice: versioning & releases (with annex lab)

Prove deep-embed in Operate with the versioning & releases slice. Use this guardrail or See example runbook. Exit metric: annex lab pass rate ≥ 0.9 with receipts linked to v2025.11-releases.

context

We need a visible, end-to-end proof that deep-embed works: a real Operate page with a runnable annex lab and tagged receipts.

decision

Ship the Operate page “Versioning & Releases” with two reversible actions and a ≤10-minute annex lab. Execute the lab in CI on PRs. Attach receipts to v2025.11-releases.

scope

In: one Operate subsection (“versioning & releases”) + its annex lab and decision entry.
Out: additional Operate subsections this cycle.

acceptance

Page uses the opener pattern above first section.
Annex lab executes locally and in CI in ≤ 10 minutes with a single command.
Frontmatter includes required fields; CTAs map to one leading and one lagging signal.
Traceability complete: pr ↔ decision/exception ↔ guardrail ↔ page ↔ signal ↔ receipt.
CI: tone at L2, traceability at L3, sanitization/red-lines at L4.
Receipts and state snapshot linked under v2025.11-releases.

receipts (expected)

leading: m-lab-pass ≥ 0.90 on pilot lab.
lagging: m-defect-rate-changed-pages does not worsen vs baseline.

freeze and exit

Freeze further Operate migrations if lab pass < 0.80 or page requires two L3 overrides. Exit after lab ≥ 0.90 and mappings pass for one release.

Pilot operate slice: versioning & releases (with annex lab) ​

context ​

decision ​

scope ​

acceptance ​

receipts (expected) ​

freeze and exit ​

links ​