Codex Review

Compatibility entry point: explicitly use Codex as the plan reviewer. The shared workflow and executable live in the sibling claudex-loop skill; install that skill alongside this one.

In Claude Code, load claudex-loop and start at mode=review, host=claude, reviewer Codex. Preserve PLAN_FILE, LOG_FILE, MAX_ROUNDS / rounds, and explicit model/effort arguments. The invocation authorizes reviewing, not implementation. Do not restart an interview when the supplied plan is already adequate.

In Codex, explain that this explicit command would use the same provider for planning and review. Use the shared claudex-loop mode=review workflow with Claude as the independent reviewer unless the user specifically requested same-provider review. If they did, perform and label that as ordinary Codex review; do not claim cross-provider approval or use the cross-provider approval runner to certify it.

Read the shared runtime reference. Resolve the runner from the installed sibling skill, not the target repository. Use the actual resolved plan path on every round and build handoff. A valid completed result and APPROVED verdict are both required; failed, empty, blocked or malformed results are not approval. The host logs findings and dispositions, and stops at the round cap.

If implementation follows, use the shared build/inspection phase. An independent plan review does not replace independent review of the final code.