codex-scenario-consistency

Align Codex/ChatGPT scenario optimizations across recall, routing, and prompts.

Updated Aug 23, 2026

Search Tags:#planning #consistency #recall #routing #codex #scenario #translator

Ready to run

npx skills add https://github.com/liuyu520/cc_source --skill codex-scenario-consistency

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: codex-scenario-consistency
Source: https://github.com/liuyu520/cc_source/tree/main/.claude/skills/codex-scenario-consistency
Command: npx skills add https://github.com/liuyu520/cc_source --skill codex-scenario-consistency

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reuse existing intent, local recall, model routing, plan mode prompt, and request translation paths to keep Codex/ChatGPT scenario optimizations consistent across all relevant entry points.

Core Features & Use Cases

Reuse existing components across intentRouter, localSearch, router, EnterPlanModeTool prompt, and translator to align behavior across Codex/ChatGPT entry points.
Apply minimal follow-through patches across multiple layers to prevent downstream drift in recall, routing, or prompting.

Quick Start

Align Codex/ChatGPT scenario changes across multiple layers with a small, targeted patch.

Frequently Asked Questions about codex-scenario-consistency

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

FAQPage Schema

How do I keep ChatGPT scenario behavior consistent across different entry points?▼

To keep ChatGPT scenario behavior consistent, you align recall, model routing, plan mode prompts, and request translation by applying minimal patches across all relevant entry points to prevent downstream drift.

What causes downstream drift in Codex recall and model routing?▼

Downstream drift in Codex recall and model routing happens when scenario optimizations are applied to one entry point without corresponding patches in intent routers, local search, and request translators.

How do I apply minimal patches to align Codex behavior across multiple layers?▼

You align Codex behavior by identifying target layers like intentRouter, localSearch, and translator, then applying the smallest feasible patches that respect existing invariants and reuse established components.

Can I reuse existing components when patching Codex plan mode prompts and routing?▼

Yes, you can and should reuse existing components when patching Codex plan mode prompts and routing to maintain system stability and respect established invariants across all patched layers.

What is the best way to validate end-to-end behavior after aligning Codex scenario optimizations?▼

The best way to validate end-to-end behavior after aligning Codex scenario optimizations is to execute real calls through the system to confirm recall, routing, and prompting consistency across all entry points.

When should I patch the request translator to prevent scenario inconsistency?▼

You should patch the request translator when scenario optimizations are introduced in intent routing or plan mode prompts, ensuring the translator applies matching behavior to prevent downstream inconsistency.