codex-architect

codex-architect is an agent for Claude Code from terrene-foundation/metis. It costs 32 tokens per session (3,837 once invoked), scanned B, original, Apache-2.0.

A specialist for designing and maintaining the files that configure coding-agent projects, such as skills, hooks, commands, and agent instructions.

In plain words
What is it for?
Use it when creating or updating .codex settings, MCP safeguards, hooks, AGENTS.md files, skills, slash commands, or related repository configuration.
Why use it?
It helps keep the agent's configuration organized and ensures the right project guidance and safeguards are in place.

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/terrene-foundation/metis/codex-architect
Clone the repo
git clone --depth 1 https://github.com/terrene-foundation/metis

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for codex-architect

README.md
[![agentmods](https://agentmods.dev/badge/agents/terrene-foundation/metis/codex-architect.svg)](https://agentmods.dev/agents/terrene-foundation/metis/codex-architect)
Your own site
<a href="https://agentmods.dev/agents/terrene-foundation/metis/codex-architect"><img src="https://agentmods.dev/badge/agents/terrene-foundation/metis/codex-architect.svg" alt="Measured on agentmods" height="20"></a>
Per session 32 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,837 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. 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 $0.00032 $0.03837
Opus 5 $0.00016 $0.01919
Sonnet 5 $0.00006 $0.00767
Haiku 4.5 $0.00003 $0.00384

Measured 3d ago against content hash 9e4cdc33e51c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

codex-architect scanned grade B with 1 finding 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 3d 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

| MCP servers | Native — `~/.codex/config.toml` `[mcp_servers.*]` blocks | [developers.openai.com/codex/config-advanced] |
.claude/agents/codex-architect.md · 167 lines

How it starts

The opening of the file, as written. The whole thing — 167 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Codex CLI Architecture Specialist

Peer to cc-architect and gemini-architect. Owns the Codex-facing substrate of every COC-enabled repo: .codex/ config tree (repo-local + user-global at ~/.codex/), .claude/codex-mcp-guard/ MCP server, .claude/hooks/*.js shim runtime, emitted AGENTS.md baseline context, and Codex-native hooks / prompts / skills.

Verified capability envelope (2026-04-22 research, Phase J1): Codex has a rich native config surface — the legacy "MCP-guard-is-the-only-enforcement-option" framing from Phase D was wrong. Codex has real hooks, skills, custom slash commands, and subagent delegation; the MCP guard remains load-bearing only for non-Bash tool enforcement (see Native Primitives table below).

Ownership Matrix (spec v6 §6.1.5)

OWNER:

  • .codex/**, .codex-plugin/**, .claude/codex-mcp-guard/** — Codex config + MCP guardrail companion
  • .claude/hooks/*.js (top-level, shim runtime) — Codex-native via .codex/hooks.json hook registration; MCP-guard fallback for non-Bash tools (see "Hooks coverage" below)
  • bin/coc-* — generated shell wrappers from .claude/wrappers/*.sh.template
  • ~/.codex/skills/<name>/SKILL.md (user-global) or .codex/skills/<name>/SKILL.md (repo-local) — native progressive-disclosure skills
  • ~/.codex/prompts/<name>.md (user-global) or .codex/prompts/<name>.md (repo-local) — custom slash commands (invoked /prompts:<name>, note the namespace prefix)

CONSUMER (read-only at emit time):

  • .claude/variants/** — slot overlays authored by cc-architect; codex-architect applies them via the emitter (Phase E4)
  • .claude/commands/** — phase commands (/analyze, /todos, etc.); codex-architect emits coc-<name> wrappers + native_primitive_override carve-outs (e.g. codex review replaces coc-review) AND emits Codex-native prompts at .codex/prompts/<name>.md
  • .claude/guides/** — copied to .codex/docs/ at sync time; no symlinks (USE template tarballs self-contained)

Primary Responsibilities

Read the full file on GitHub · 167 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. 3d ago First seen · 167 lines · 32 tokens per session scan B 9e4cdc33e51c

Subscribe to this mod's changes

codex-architect is an agent published in the GitHub repository terrene-foundation/metis (2 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 32 tokens to every session and 3,837 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.