codex-rescue

Delegate stuck Claude Code problems to Codex CLI for second-opinion resolution.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/michaelkd01/dic-skills --skill codex-rescue
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex-rescue
Source: https://github.com/michaelkd01/dic-skills/tree/main/codex-rescue
Command: npx skills add https://github.com/michaelkd01/dic-skills --skill codex-rescue

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Claude Code sessions sometimes loop on the same issue, exhaust their iteration quota, or reveal blind spots that stall progress in manual workflows. This guide explains how to identify those trapped states, why rescue is a second-opinion intervention, and how to safely hand off the specific failing piece to Codex while keeping the manual session in control.

Core Features & Use Cases

  • Command-driven delegation: Use /codex:rescue with --background, --wait, --resume, --fresh, --model, and --effort to describe the precise failure and request the right level of investigation or fix.
  • Lifecycle tracking: Monitor and manage jobs with /codex:status, /codex:result, and /codex:cancel, remembering that the rescue runs on the same checkout and that you should review Codex output before resuming Claude Code.
  • Use Case guidance: Follow the three-strike rule, iteration budget guardrail, targeted investigations, or model-switch patterns, and optionally set project-level defaults in .codex/config.toml (or user defaults) to preselect models, effort levels, and behavior.

Quick Start

Describe the stuck Claude Code work to Codex and request a rescue with the appropriate flags so it can investigate or continue while you monitor the output before resuming the session.

Frequently Asked Questions about codex-rescue

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

FAQPage Schema
How do I get a second opinion when my Claude Code session is stuck?

To get a second opinion for a stuck Claude Code session, use the /codex:rescue command to delegate the failing piece to Codex. Codex investigates or continues the work while you monitor the output before resuming your session.

What is the three-strike rule for escalating debugging tasks to another model?

The three-strike rule is a use-case pattern for manual Claude Code workflows: after repeated failures or exhausted iterations, you escalate the specific failing piece to Codex as a second-opinion intervention rather than continuing to loop.

Do I need the Codex CLI installed to rescue a Claude Code task?

Yes, you need the Codex CLI installed and an authenticated plugin to execute rescue jobs. The rescue runs on the same checkout, so you must review Codex output before resuming your Claude Code session.

How do I monitor a background rescue job delegated to Codex?

To monitor a background rescue job, use the /codex:status command to track its lifecycle and /codex:result to retrieve the output. Use /codex:cancel to stop a job if needed.

Can I configure default models and effort levels for Codex rescue tasks?

Yes, you can set project-level defaults in the .codex/config.toml file or user defaults to preselect models, effort levels, and behavior. You can also override these per rescue using the --model and --effort flags.

When should I not use a model handoff for a Claude Code debugging loop?

You should not use a model handoff for simple tasks that have not hit repeated failures or exhausted iteration quotas. Rescue is a targeted second-opinion intervention for stalled manual workflows needing investigation before executing fixes.