[ElliotDrel]

What Changed and Why

Issue: Better Title assumed a Claude Code session and could not route title operations from Codex or explain an unsupported host.

Why this change was necessary: Session identifiers and rename mechanisms differ between hosts. Nested CLI sessions can expose both environments, and a stale identifier must not rename another chat.

New status quo: Better Title 2.0.2 detects the host, reads its current title, and prints the matching rename command. Claude retains append-only transcript updates; Codex uses thread/read and thread/name/set through its app-server. Unknown, ambiguous, missing, conflicting, or stale session identifiers stop with diagnostics. Explicit host selection keeps the session safeguard. Startup failures remain visible to the agent, and Codex has an explicit-invocation policy. Unicode and literal whitespace survive the transport. Installer/prefix work is excluded.

Verification

Codex verified rebased commit 939495c on main 18c7478 (1.1.1), preserving the prompt audit and current feedback template. Local verification used Windows with Codex CLI 0.162.1:

• 29 isolated routing/transport tests pass locally, including a permission-denied startup case. All six CI jobs pass at 939495c: title routing on Windows/Linux and installer tests on both platforms with Node 18.0.0 and 24. Prefix-safety tests, installer edge cases, and repository gates also pass on the rebased commit.
• 196 existing Python tests, migration smoke tests, and installer edge cases passed.
• Skill names, paths, versions, docs, feedback, Bash/Node syntax, and diff whitespace checks passed.
• Real Codex app-server read/write/read passed in a disposable CODEX_HOME, without a model turn or existing user chat.

Claude used temporary transcript fixtures. Live Claude UI refresh and a real daemon proxy were not tested. No live skill install or npm release was performed. No lint/typecheck/build/formatter tooling is configured.

Configuration

Codex requires host-tool approval for startup and rename when its home is outside sandbox write roots. The skill requests that approval; it stops if approval is denied or unavailable. The real app-server smoke test used an approved escalated shell.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-19T17:22:33.042977Z	fb17482	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: fb17482d7c

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 72380524-4f99-46a1-8547-af730746319f

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}