Run Codex Lab dogfood readiness sweep #52

Current Status

State: Complete. Dogfood readiness sweep is green as of 2026-06-12 on plan/dogfood-readiness-sweep from current main (ba00aca888891935c7f3f8cc8370e985f54e22b9). No repo file changes were needed.

Go/no-go: GO for local daily-driver dogfooding of the covered MVP spine.

Evidence:

• just exec-harness-test passed all 5 scenarios and wrote .tmp/codex-exec-harness/report.json with passed: true.
• skills-cache-continuity passed inside the full harness with cached-token proof on turn two (cached_input_tokens: 1500, total_tokens: 3165).
• python3 -m unittest discover -s tools/codex-exec-harness -p 'test_harness.py' passed 31/31 Auto Review proof-loop tests.
• just write-app-server-schema completed and schema paths stayed clean.
• just test -p codex-app-server-protocol passed 226/226.
• Targeted app-server Desktop compatibility tests from codex-rs/app-server/README.md passed 11/11.
• just test -p codex-cli strict_config_rejects_unknown_config_fields_for_app_server passed 1/1.
• Docs/stale-promise sweep found no blockers and no reintroduced retired MVP surfaces (/auto, code exec --auto, Auto Drive coordinator/cards/settings, app-server v1 conversation resurrection).

Blockers: none for local dogfooding of the covered spine.

Next action: start #45, the Code Bridge protocol/trust/payload contract, before service skeleton work in #46.

Finish Line

Codex Lab has a current, evidence-backed dogfood readiness result that says whether the covered MVP spine is green enough to use as the daily driver before Code Bridge implementation starts.

Acceptance Criteria

• Run or explicitly account for the current exec-harness gate.
• Run or explicitly account for the skills/cache continuity scenario.
• Run or explicitly account for Auto Review proof-loop coverage.
• Run or explicitly account for app-server/Desktop compatibility gates from codex-rs/app-server/README.md.
• Review top-level docs/planning prose for stale claims about unbuilt Every Code parity.
• Record dogfood go/no-go result and blockers in this issue and Codex Lab MVP dogfood plan #28.

Relationships

Child of #28. Follows #30. Should complete before starting #45/#46 unless the user explicitly chooses to begin Code Bridge first.