Codex

Bridge Claude with OpenAI's Codex CLI. Access GPT-5.6 Sol for code generation, analysis, and review with automatic fallback to GPT-5.6 Terra on quota limits.

Best for: targeted code reasoning, architecture critique, and security review of specific files — the default workhorse reviewer. Not for: whole-repository reads beyond its context window (use Gemini or Antigravity), or fully offline/air-gapped use (it's a hosted model — use Ollama).

Installation

Run in your terminal:

# Project scope (current project only)
claude mcp add codex-cli -- npx -y ask-codex-mcp

# User scope (all projects)
claude mcp add --scope user codex-cli -- npx -y ask-codex-mcp

Or install as a plugin (adds slash commands like /multi-review, /brainstorm, /compare, plus reviewer subagents and the opt-in continuous codex-pair review hook):

/plugin marketplace add Lykhoyda/ask-llm
/plugin install ask-llm@ask-llm-plugins

Or install globally:

npm install -g ask-codex-mcp

Prerequisites

1. Node.js v20.0.0 or higher
2. Codex CLI installed and authenticated

Tools

Tool	Purpose
ask-codex	Send prompts to Codex CLI. Optional reasoningEffort and sessionId for multi-turn — the latter maps to Codex's native thread_id and uses codex exec resume <id> under the hood (ADR-058)
get-usage-stats	Per-session token totals + breakdowns. In-memory (ADR-054)
ping	Fast connection test to verify MCP setup

ask-codex returns both human-readable text and a structured AskResponse (provider, response, model, sessionId, usage) via MCP outputSchema. The Thread ID returned in the response footer is the same value as structuredContent.sessionId — pass it back as sessionId to continue the conversation.

Models

• Default: gpt-5.6-sol (flagship capability)
• Fallback: gpt-5.6-terra (balanced tier, automatic on quota errors per ADR-028)

Key Features

• GPT-5.6 Sol access via the official Codex CLI
• Reasoning control — ordinary calls default to medium; /codex-review and /brainstorm use high; direct calls can request low, medium, high, xhigh, or max
• Native session continuity — sessionId parameter maps to Codex's thread_id; codex exec resume <id> is used internally for follow-up turns (zero replay cost — Codex retains state)
• Non-interactive sandbox — codex exec --sandbox workspace-write so Codex never hangs on approval prompts in MCP subprocess context (--full-auto was removed upstream in codex 0.128; exec is non-interactive by definition — ADR-075)
• JSONL output parsing for structured responses + token usage
• Automatic quota fallback from GPT-5.6 Sol to Terra
• Structured AskResponse via outputSchema for programmatic clients
• Standard MCP transport works with 40+ clients

npm

• Package: ask-codex-mcp
• Binary: ask-codex-mcp