Codex

OpenAI Codex extends through MCP servers declared in ~/.codex/config.toml and AGENTS.md instructions. Wire up the MCP server with Smithers:

bunx smithers-orchestrator mcp add --agent codex

That registers the MCP server in Codex’s config. Restart Codex and it can drive workflows.

Model tiers

Smithers pins each Codex worker to the tier that matches its role:

Work	Model
Planning, orchestration, review, and the hardest reasoning	`gpt-5.6-sol`
Mid-tier judgment, validation, or tool-heavy work that needs extra checking	`gpt-5.6-terra`
Research, implementation, ordinary tool use, UI, and routine-to-substantial execution	`gpt-5.6-luna`

Use the exact tier IDs rather than the floating gpt-5.6 alias. Start with Luna for most research, implementation, ordinary tool use, and substantial execution; size alone does not require Sol. Escalate to Terra for stronger validation or structured tool-heavy judgment, and to Sol for ambiguity, high-stakes decisions, novel architecture, orchestration, final review, or repeated failure. When usable Codex authentication is present, generated Smithers workflows put these Codex workers first. Other adapters remain later runtime fallbacks and run only when the Codex attempt is unavailable or fails. See SOTA role defaults.

Smithers docs in Codex

Smithers does not currently auto-install a Codex skill, and bunx smithers-orchestrator skills add has no supported --agent codex target filter. For Codex, keep the repo-level AGENTS.md guidance below and use the packaged docs commands when Codex needs the full API:

bunx smithers-orchestrator docs-full
bunx smithers-orchestrator ask "How do I run and watch a workflow?"

Register the MCP server

mcp add writes the entry for you. The native config is a TOML table in ~/.codex/config.toml:

[mcp_servers.smithers]
command = "bunx"
args = ["smithers-orchestrator", "--mcp"]

Or use the Codex CLI:

codex mcp add smithers -- bunx smithers-orchestrator --mcp

Codex launches the server at session start and the semantic tools (list_workflows, run_workflow, watch_run, resolve_approval, …) appear in the /mcp list. Full reference: MCP Server.

Standing instructions

Codex reads AGENTS.md from the repo root down to the working directory. Add a section so Codex knows to reach for Smithers:

## Orchestration
Use Smithers for durable, multi-step, or human-in-the-loop agent work.
Run `bunx smithers-orchestrator starters` to see ready-made workflows,
then `bunx smithers-orchestrator workflow run <name> --prompt "…"`.

This same AGENTS.md is read by Cursor, Copilot, Hermes, and Pi; one file covers several agents.

Smithers runs Codex too

In the other direction, Smithers spawns Codex as a worker via CodexAgent (codex exec with a streamed JSON protocol and native thread hijack):

import { CodexAgent } from "smithers-orchestrator";

const implementer = new CodexAgent({
  model: "gpt-5.6-luna",
  config: { model_reasoning_effort: "medium" },
  sandbox: "workspace-write",
});

See CLI Agents → Codex.

Set up in your harness

Claude Code

Start

Articles

Learn

Build Workflows

Run and Operate

Workflow Pack

Components

Integrations

Agent Support

Examples

Contributing

Model tiers

Smithers docs in Codex

Register the MCP server

Standing instructions

Smithers runs Codex too

​Model tiers

​Smithers docs in Codex

​Register the MCP server

​Standing instructions

​Smithers runs Codex too

Model tiers

Smithers docs in Codex

Register the MCP server

Standing instructions

Smithers runs Codex too