execute-codex

A command that routes coding work between different AI models based on the task size, then has another model review the result. It follows a staged workflow for planning, implementation, and approval.

In plain words
What is it for?
Use it to execute a prepared coding plan or task, with larger work assigned to Codex, smaller work to Claude, and review handled by Sonnet.
Why use it?
It organizes model roles and adds a review step before changes are delivered. It also defines rules for handling missing context and protecting secrets.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/ching-kuo/claude-codex/execute-codex
Clone the repo
git clone --depth 1 https://github.com/ching-kuo/claude-codex
Per session 14 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,065 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00014 $0.02065
Opus 5 $0.00007 $0.01033
Sonnet 5 $0.00003 $0.00413
Haiku 4.5 $0.00001 $0.00206

Measured 2d ago against content hash 54924aad0828, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

execute-codex scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/execute-codex.md · 159 lines

How it starts

The opening of the file, as written. The whole thing — 159 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Deprecated: This command has been converted to a skill (/execute-codex). The skill version is recommended for new usage and supports eval-based testing. This command is retained for backward compatibility (model pinning, tool restrictions).

Execute-Codex — Smart Routing: Claude (small) / Codex (large)

$ARGUMENTS


Core Protocols

  • Code Sovereignty: Claude is the final authority — review and approve all changes before delivery
  • Stop-Loss: Do not proceed to next phase until current phase output is validated
  • Language: Use English when calling tools/models; communicate with user in their language
  • Context Sanitization: Never pass .env, secrets, tokens, API keys, or credentials to any external agent or MCP. Exclude files matching .env*, *secret*, *credential*, *.pem, *.key. Redact inline secrets before sending.

Execution Workflow

Phase 0: Read Plan

  1. If argument is a file path, read it and extract: task description, implementation steps, key files
  2. If no plan file, use the argument as the task description directly
  3. Confirm with user before proceeding if key context is missing

Phase 1: Context Retrieval

Read key files using Read, Glob, Grep. Confirm complete context before proceeding.

Phase 2: Route by Change Size

Classify the task before doing any implementation:

Small change — ALL of the following must be true:

  • Touches ≤ 2 files
  • Estimated diff ≤ 30 lines
  • No new abstractions, modules, or significant logic (e.g. config edits, wording fixes, minor additions, single-function tweaks)

Large change — any of the following:

  • Touches 3+ files
  • Estimated diff > 30 lines
  • Introduces new logic, new files, refactoring, or cross-cutting concerns

Announce your routing decision before implementing (e.g. "Small change — implementing directly" or "Large change — routing to Codex").


Route A: Small Change — Claude Implements Directly

If the plan has 3+ implementation tasks, dispatch each task to a subagent (sequentially):

  • Before each task: record $TASK_SHA via git rev-parse HEAD
  • For each task: launch a general-purpose Task agent with the task description, sanitized context files, and instruction: "Implement only this task. Run available lint/tests scoped to changed files. Report: files changed, verification result, any blockers."
  • After each subagent: run git diff $TASK_SHA (scoped to this task only) to verify changes, run scoped lint/tests to confirm no regressions
  • If the subagent reports a blocker: fix directly with Edit/Write or re-dispatch with failure output

Read the full file on GitHub · 159 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 159 lines · 14 tokens per session scan A 54924aad0828

Subscribe to this mod's changes

execute-codex is a command published in the GitHub repository ching-kuo/claude-codex (23 stars, last pushed 5mo ago), licensed MIT. It adds 14 tokens to every session and 2,065 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.