mcpField
O
oracle-codex

oracle-codex

github.com/paulrbergAgent workflows102190 installsUpdated today
Repository

About oracle-codex

A skill for using OpenAI Codex CLI as a read-only source for planning, code review, and analysis. It requires a preflight check and routes invocations through a wrapper, while prohibiting implementation changes.

Key features

  1. 01Read-only Codex analysis
  2. 02Configurable reasoning effort
  3. 03Prerequisite check before use
  4. 04Wrapper for Codex invocations

Use cases

  1. 01Reviewing a code change
  2. 02Planning an implementation
  3. 03Analyzing a codebase without edits

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/paulrberg/agent-skills --skill oracle-codex

After installing, type/oracle-codexin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

references/
codex-flags.md
scripts/
check-codex.sh
run-codex-exec.sh

Details

From the project README

Codex Oracle

Use OpenAI Codex CLI as a read-only oracle — planning, review, and analysis only. Codex provides its perspective; you synthesize and present results to the user.

Sandbox is always read-only. Codex must never implement changes.

Arguments

Parse $ARGUMENTS for:

  • query — the main question or task (everything not a flag). Required — if empty, tell the user to provide a query and stop.
  • --reasoning <level> — override reasoning effort (low, medium, high, xhigh). Optional; default is xhigh.

Prerequisites

Run the check script before any Codex invocation:

scripts/check-codex.sh

If it exits non-zero, display the error and stop. Use the wrapper for all codex exec calls:

scripts/run-codex-exec.sh

Configuration

SettingDefaultOverride
Modelgpt-5.5Allowlist only (see references/codex-flags.md)
Reasoningxhigh--reasoning <level> or user prose
Sandboxread-onlyNot overridable
Reasoning Effort
ComplexityEffortTimeoutCriteria
Simplelow300000ms<3 files, quick question
Moderatemedium300000ms3–10 files, focused analysis
Complexhigh600000msMulti-module, architectural thinking
Maximumxhigh600000msFull codebase, critical decisions

For xhigh tasks that may exceed 10 minutes, use run_in_background: true on the Bash tool and set CODEX_OUTPUT so you can read the output later.

See references/codex-flags.md for full flag documentation.

Workflow

1. Parse and Validate
  1. Parse $ARGUMENTS for query and --reasoning
  2. Run scripts/check-codex.sh — abort on failure
  3. Use xhigh reasoning effort unless --reasoning overrides it
2. Construct Prompt

Build a focused prompt from the user's query and any relevant context (diffs, file contents, prior conversation). Keep it direct — state what you want Codex to analyze and what kind of output you need. Do not implement; request analysis and recommendations only.

3. Execute

Invoke via the wrapper with HEREDOC. Set the Bash tool timeout per the reasoning effort table above.

EFFORT="<effort>" \
CODEX_OUTPUT="/tmp/codex-${RANDOM}${RANDOM}.txt" \
scripts/run-codex-exec.sh <<'EOF'
[constructed prompt]
EOF

For xhigh, consider run_in_background: true on the Bash tool call, then read CODEX_OUTPUT when done.

4. Present Results

Read the output file and present with attribution:

## Codex Analysis

[Codex output — summarize if >200 lines]

---
Model: gpt-5.5 | Reasoning: [effort level]

Synthesize key insights and actionable items for the user.

Comments

Markdown supported