codex-advisor

Queries OpenAI Codex CLI in non-interactive mode for plan reviews, code reviews, and architecture decisions.

1|Updated May 6, 2026
One-click install
npx skills add https://github.com/surfingalien/FinSurfing --skill codex-advisor-surfingalien
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex-advisor
Source: https://github.com/surfingalien/FinSurfing/tree/main/.claude/skills/codex-advisor
Command: npx skills add https://github.com/surfingalien/FinSurfing --skill codex-advisor-surfingalien

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @openai/codex.

What problem does it solve? When you are stuck on a hard problem, unsure about an implementation plan, or want a second opinion on security-sensitive code, this Skill lets you consult OpenAI's Codex CLI as an external advisor without leaving your workflow. ## Core Features & Use Cases - Plan Review: Validate implementation plans against your actual codebase before writing code, catching missing edge cases and security concerns. - Code Review: Pipe git diffs into Codex for line-by-line feedback on bugs, vulnerabilities, and performance issues. - Architecture Decisions: Compare design trade-offs with codebase-aware reasoning using models like gpt-5.2 and gpt-5.2-codex. - Use Case: Before committing a large authentication refactor, run git diff --staged through Codex with xhigh reasoning to catch logic errors and missing error handling. ## Quick Start Ask the advisor to review your staged changes by piping the git diff into codex exec with the gpt-5.2-codex model and xhigh reasoning effort.

Frequently Asked Questions about codex-advisor

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

FAQPage Schema
How do I review a git diff with OpenAI Codex CLI?

Pipe the diff into codex exec with a code-focused model: git diff --staged | codex exec -m gpt-5.2-codex -c model_reasoning_effort="xhigh" "Review these changes". Always use exec mode since interactive mode cannot accept piped input.

Which Codex model should I use for code review?

Use gpt-5.2-codex for code reviews and debugging tasks. Use gpt-5.2 for general plan reviews and architecture questions, gpt-5.1-codex-max for long multi-file migrations, and gpt-5.1-codex-mini when cost matters.

How do I authenticate Codex CLI in CI environments?

Set the CODEX_API_KEY environment variable with an API key for CI pipelines. For local use, run codex --login once interactively or export OPENAI_API_KEY.

Why does Codex CLI fail with 'stdin is not a terminal'?

This error occurs when piping data into interactive mode. Use codex exec instead of plain codex for all non-interactive and piped usage, such as git diff | codex exec -m gpt-5.2 "Review this".

When should I not use an AI advisor for code review?

Skip it for trivial changes like typos, formatting, or obvious bug fixes where the solution is clear. Also avoid sending proprietary code, credentials, or PII to external servers without checking your organization's policies.