third_party_review

Automate iterative Codex code review feedback on PRs until approval.

1|Updated Oct 19, 2025
One-click install
npx skills add https://github.com/acartine/osterman --skill third-party-review
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: third_party_review
Source: https://github.com/acartine/osterman/tree/main/skills/third_party_review
Command: npx skills add https://github.com/acartine/osterman --skill third-party-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reduces operator thrashing by delegating code review to Codex, enabling automated, iterative feedback on PRs and branches until approval.

Core Features & Use Cases

  • Automated third-party code review feedback via Codex
  • Iterative review loop until APPROVED or NEEDS_WORK
  • Post reviews as PR comments and continue the workflow
  • Safeguards: max iterations and operator notifications

Quick Start

Trigger the third_party_review on your current branch to start automated Codex review feedback.

Frequently Asked Questions about third_party_review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate code review feedback on a pull request using Codex?

Automated code review feedback on a pull request is generated by triggering an iterative loop with Codex until the branch receives an APPROVED status, posting the critique directly as PR comments.

How does iterative code review feedback work until a PR is approved?

Iterative code review feedback works by applying Codex-based critique to a PR repeatedly, posting comments, and continuing the workflow until the review loop reaches APPROVED or NEEDS_WORK within a maximum iteration limit.

What inputs do I need to run automated Codex reviews on a git branch?

Running automated Codex reviews on a git branch requires a pr_number and repository inputs, alongside integration with codex, gh, and git to post review content as PR comments.

What is the best way to stop operator thrashing during pull request reviews?

The best way to stop operator thrashing during pull request reviews is delegating the independent critique to Codex, which enforces a maximum iteration limit and handles iterative feedback until consensus is reached.

What happens when automated code review iterations hit the maximum limit?

When automated code review iterations hit the maximum limit, the iterative review loop terminates and triggers operator notifications to prevent infinite cycles and ensure manual intervention.

When do I need an independent automated critique on a PR or branch?

You need an independent automated critique on a PR or branch when you want to reduce operator thrashing and enforce iterative code review feedback via Codex before merging changes.