Find the Best Cosmetic Hospitals

Explore trusted cosmetic hospitals and make a confident choice for your transformation.

“Invest in yourself — your confidence is always worth it.”

Explore Cosmetic Hospitals

Start your journey today — compare options in one place.

Private Self-Review of GitHub PRs with Codex (Without Posting Comments)

Private Self-Review of GitHub PRs with Codex (Without Posting Comments)

If you like Codex as a reviewer but don’t want its feedback to show up publicly on your pull requests, the trick is simple:

  • Use the GitHub integration only when you want a visible PR review.
  • Use Codex CLI (local) for private self-review—nothing gets posted back to GitHub.

This tutorial shows exactly how to set that up.


Why this matters

When Codex is connected to GitHub PRs, its review output appears in the PR like any other review comment. Anyone with access to the PR can see it.

So if your goal is:

✅ “Review only when I ask”
✅ “Keep the feedback private for my own self-review”

…then local review via Codex CLI is the right workflow.


Part 1: Stop Codex from auto-reviewing PRs

In Codex → Settings → Code review:

  1. Keep Personal auto review preferences = OFF
  2. Don’t enable repository-level automatic review triggers (like “review on PR open”)

This ensures Codex won’t review every PR automatically.

Result: Codex reviews happen only when you explicitly request them.


Part 2: Understand “Will Codex review the entire codebase?”

No. Codex CLI review is usually focused on your diff:

  • your PR branch vs a base branch (ex: main)
  • a single commit
  • only uncommitted changes

Codex may open surrounding files for context, but it is not doing a full repo-wide audit unless you explicitly ask for that kind of analysis.


Part 3: Private self-review of a PR using Codex CLI

Step 1 — Check out the PR branch locally

If you use GitHub CLI:

gh pr checkout <PR_NUMBER>
Code language: HTML, XML (xml)

Or with git:

git fetch origin pull/<PR_NUMBER>/head:pr-<PR_NUMBER>
git checkout pr-<PR_NUMBER>
Code language: HTML, XML (xml)

Step 2 — Run Codex CLI in your repo

cd /path/to/your/repo
codex

Step 3 — Run a private review

Inside Codex CLI:

/review

Then choose what you want to review:

  • Branch vs base (recommended): review your PR branch compared to main
  • Single commit: useful if your PR is large—review in chunks
  • Uncommitted changes: great before you commit or push

Step 4 — Confirm what’s being reviewed (optional)

To see the exact diff Codex is looking at:

/diff

That’s it.
You now get review feedback privately in your terminal/IDE and nothing is posted to the PR.


Best practice workflow (simple and effective)

  1. Before pushing: run /review on uncommitted or staged changes
  2. Before opening PR: run /review against main
  3. Only when needed: use @codex review in GitHub to get a team-visible review

This gives you both:

  • private self-review whenever you want
  • public review only when you choose

FAQ

1) Can Codex review my PR privately inside GitHub without others seeing it?

No. If Codex posts a review to the PR via the GitHub integration, it’s visible to anyone who can access that PR. For private feedback, run the review locally using Codex CLI (or your IDE) instead.

Yes — it can “cost” in the sense that it consumes your Codex allowance (quota), and may consume paid credits if you go past the included limits.

If you’re using Codex with your ChatGPT login (Plus / Pro / Business / Enterprise / Edu)

  • No extra charge up to your plan’s included limits — Codex is included with these plans. (OpenAI Developers)
  • Your PR self-review (via Codex CLI /review) will consume your local-message / local-task usage window. (OpenAI Developers)
  • If you hit the included limits, you can extend usage by buying ChatGPT credits (so then it does cost). (OpenAI Developers)

If you trigger review in GitHub (@codex review)

  • That counts as Code Review usage (cloud/GitHub), and it’s the type that’s explicitly “charged” in the rate card / credits model. (OpenAI Help Center)

If you authenticate Codex with an API key

  • Then it’s pay-as-you-go API billing (“pay only for the tokens Codex uses”), separate from your ChatGPT subscription. (OpenAI Developers)

How to check what you’re spending

  • In Codex: usage dashboard (and in CLI you can use /status to see remaining limits). (OpenAI Developers)

If you tell me which plan you’re on (Plus/Pro/Business/Enterprise) and whether you’re using GitHub @codex review or CLI /review, I can tell you exactly which quota bucket you’re consuming.

Find Trusted Cardiac Hospitals

Compare heart hospitals by city and services — all in one place.

Explore Hospitals
Subscribe
Notify of
guest
0 Comments
Newest
Oldest Most Voted
Inline Feedbacks
View all comments

Certification Courses

DevOpsSchool has introduced a series of professional certification courses designed to enhance your skills and expertise in cutting-edge technologies and methodologies. Whether you are aiming to excel in development, security, or operations, these certifications provide a comprehensive learning experience. Explore the following programs:

DevOps Certification, SRE Certification, and DevSecOps Certification by DevOpsSchool

Explore our DevOps Certification, SRE Certification, and DevSecOps Certification programs at DevOpsSchool. Gain the expertise needed to excel in your career with hands-on training and globally recognized certifications.

0
Would love your thoughts, please comment.x
()
x