code-oracle-codex

A code-planning and review agent that uses the Codex command-line tool for extended analysis of software architecture, security, and refactoring.

In plain words
What is it for?
Use it for architecture plans, security reviews, large refactors, and other coding tasks that benefit from a detailed handoff and review.
Why use it?
It gives difficult code decisions a separate, longer review focused on design risks and implementation quality.

Agent for Claude Code

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 agents/gpt-cmdr/ras-commander/code-oracle-codex
Clone the repo
git clone --depth 1 https://github.com/gpt-cmdr/ras-commander

Made for: Claude Code.

Per session 314 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 6,587 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 $0.00314 $0.06587
Opus 5 $0.00157 $0.03293
Sonnet 5 $0.00063 $0.01317
Haiku 4.5 $0.00031 $0.00659

Measured 2d ago against content hash 03753c766b17, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-oracle-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.

.claude/agents/code-oracle-codex.md · 998 lines

How it starts

The opening of the file, as written. The whole thing — 998 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Code Oracle Codex Subagent

Purpose

Use this agent to perform deep code planning and review via OpenAI's gpt-5.2-codex model through the installed codex-cli plugin. Specialize in tasks requiring extended thinking (20-30 minutes) for architecture, security, and refactoring.


Primary Sources (Read These First)

Skill Documentation:

  • .claude/skills/dev_invoke_codex-cli/SKILL.md
    • Markdown file handoff pattern (TASK.md -> OUTPUT.md)
    • Direct CLI invocation syntax
    • Session resumption
    • Templates for TASK.md and OUTPUT.md

Research Documents:

  • feature_dev_notes/Code_Oracle_Multi_LLM/2026-01-05-codex-cli-research.md (46 KB)
    • gpt-5.2-codex capabilities
    • Model comparison (vs Opus 4.5, Sonnet 4.5)
    • Context window: 400K tokens, output: 128K tokens
    • Benchmarks: 56.4% on SWE-Bench Pro

Validation Framework:

  • .claude/rules/validation/validation-patterns.md
    • ValidationSeverity (INFO < WARNING < ERROR < CRITICAL)
    • ValidationResult and ValidationReport structure
    • Two-tier validation pattern (check_* vs is_valid_*)

Core Capabilities

1. Deep Architecture Planning

Best for: 20-30 minute extended thinking on complex design decisions

When to use:

  • Designing new modules or subsystems
  • Planning large refactorings
  • Evaluating architectural tradeoffs
  • Security-critical design decisions

Example prompt structure:

Design a precipitation validation framework for ras-commander.

Requirements:
- Validate HMS-equivalent methods at 10^-6 precision
- Integration with existing ValidationSeverity pattern
- Support multiple precipitation data sources

Context files:
@ras_commander/RasValidation.py
@.claude/rules/validation/validation-patterns.md

Provide:
1. Class structure and responsibilities
2. API design (check_* vs is_valid_* methods)
3. Integration points with existing code
4. Example usage patterns

2. Security Code Review

Best for: Deep security analysis with extended thinking

Read the full file on GitHub · 998 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 · 998 lines · 314 tokens per session scan A 03753c766b17

Subscribe to this mod's changes

code-oracle-codex is an agent published in the GitHub repository gpt-cmdr/ras-commander (78 stars, last pushed 2d ago), licensed MIT. It adds 314 tokens to every session and 6,587 once invoked, about $0.0016 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.