Chief + Codex: Better Together

At a glance

What is Chief?

Chief is an AI-native agent manager. It sits above coding agents like Codex and handles everything between "here's what I want" and "here's the shipped result" — planning what to build, figuring out what comes next, dispatching agents across your portfolio, and reviewing output before it ships.

You already use Codex to implement features. Chief makes Codex productive at scale — across every project you manage, with the right agent for each task, executing complex goals end-to-end.

Chief uses a BYOA (bring-your-own-agent) model. You connect your existing Codex subscription — no new agent costs, no migration.

What Chief adds on top of Codex

1. Plan and execute complex goals in one shot

Codex's Goal Mode drives toward an objective autonomously — powerful for exploration and greenfield work. But it's unstructured: you don't see the dependency graph, can't inspect checkpoints along the way, and don't get acceptance criteria for each step.

Chief provides structured one-shot execution. Give it "implement Stripe billing with usage metering and team plans" and Chief decomposes it into dependency-aware milestones — the schema migration before the API endpoints, the webhook handler before the billing page. Each milestone has acceptance criteria and a review gate. You approve the plan once; Chief executes the entire sequence with checkpoints you can inspect.

The difference: Goal Mode says "go build this" and you see the result. Chief says "here's the plan — approve it, and I'll execute and verify each step."

2. Proactive management that figures out what's next

Codex waits for tasks. Each task is an island — you decide what to work on, queue it, and Codex executes. Between tasks, Codex is idle until you return.

Chief is proactive. It analyzes project state, understands what's been completed, what's blocked, and what's ready to start — then proposes the next milestones. When morning milestones ship, Chief has already figured out the afternoon's priorities. You make decisions; Chief drives the pace.

3. Scale agents across your entire portfolio

Running Codex across five repos gives you five independent agents with no awareness of each other. Each task is isolated — no portfolio view, no conflict detection, no cross-project dependency tracking.

Chief manages agents across all your projects simultaneously. Dispatch Codex to three repos, Claude Code to two more, and see all progress in one dashboard. Chief tracks what's in flight everywhere, and sequences cross-project dependencies. Your output multiplies with each project you add — without multiplying your management overhead.

4. Mix the right provider for each task

Codex runs only OpenAI models. If Claude handles a specific task better, or you want to use the best model for each type of work, Codex can't route there.

Chief's BYOA model lets you mix providers per task. Use Codex where GPT-5.5 shines — rapid prototyping, broad language support — and Claude Code where Opus excels — complex refactors, nuanced architecture decisions. Chief manages both from a single interface, routing the right agent to the right milestone. You get the best of every provider instead of being locked into one.

5. Team-wide visibility

When multiple team members use Codex on the same codebase, there's no shared view of what's in flight. Coordination falls on Slack messages and "don't touch auth today."

Chief gives the whole team a unified dashboard showing all in-progress milestones across every team member and project. Everyone sees what's in flight, what's completed, and what's queued — replacing ad-hoc Slack coordination with real-time visibility.

6. Quality gates at scale

Every Codex task ends with the developer reviewing output. At low parallelism this works. At high parallelism — five projects, multiple agents — you become the review bottleneck.

Chief reviews agent output against acceptance criteria before surfacing results to you. It checks for regressions, validates the implementation, and flags issues. You handle the decisions that need human judgment, not the first-pass quality checks.

When to add Chief

Add Chief to your Codex workflow when:

• You want to scale across projects. Chief lets you manage agents across your entire portfolio from one dashboard — not five independent task queues.
• You want one-shot execution of complex goals. Give Chief a goal, approve the plan, and it executes the full milestone sequence with checkpoints and review gates.
• You want proactive management. Chief figures out what to work on next based on project state — you stop being the bottleneck between tasks.
• You want to mix providers. Route Codex and Claude Code to the tasks each handles best, managed from a single interface.
• You're on a team. Unified visibility shows what every agent is doing.
• You want quality gates at scale. Chief reviews agent output against acceptance criteria before it hits your inbox.

Chief doesn't replace Codex — it turns Codex from a tool you operate into an agent that's managed as part of a portfolio-wide operation.

Getting started in 5 minutes

Already using Codex? Adding Chief takes minutes.

1. Sign up for Chief — free trial, no credit card.
2. Connect your Codex subscription — Chief uses your existing OpenAI account via BYOA. No additional agent costs.
3. Import a project — give Chief a GitHub repo URL. It builds project context automatically.
4. Set a goal — describe what you want accomplished in plain language.
5. Approve the plan — Chief proposes milestones with acceptance criteria. Review and approve.
6. Watch it execute — Chief dispatches Codex with the right context, runs the full plan, and reviews the output.

Your Codex workflow stays the same — it just gains a manager.

FAQ

Is Chief a replacement for Codex?

No. Chief manages Codex; it doesn't replace it. Codex does the coding. Chief does the planning, dispatching, and reviewing. They're different layers — manager and engineer.

Do I need a separate Codex subscription?

You bring your existing Codex / OpenAI subscription. Chief uses a BYOA model — no additional agent costs from Chief's side.

Can Chief manage Codex alongside other agents?

Yes. Chief's BYOA model supports multiple agents. Use Codex for some tasks and Claude Code for others — Chief dispatches the right agent for each milestone from a single interface. That's the core value of BYOA: mix and match the best provider for each job.

How is Chief's planning different from Codex's Goal Mode?

Goal Mode drives toward an objective autonomously — it works for hours making its own decisions, and you see the result. Chief provides structured decomposition: dependency-aware milestones with acceptance criteria, sequenced execution, and review gates at each checkpoint. Goal Mode is autonomous execution. Chief is managed execution with visibility into every step.

How does Chief decide what to work on next?

Chief maintains persistent project context — completed milestones, in-flight work, project goals, codebase state. It uses this context to proactively propose the next milestones. You approve or redirect; Chief drives the pace.

What does "one-shot execution" actually mean?

You give Chief a complex goal — "add team billing with Stripe, usage metering, and an admin dashboard." Chief decomposes it into dependency-aware milestones, sequences them, and executes the entire plan. Each milestone is reviewed against acceptance criteria before the next begins. You approve once at the plan level, not at every step.

Can I still use Codex directly for ad-hoc tasks?

Yes. Chief manages structured, planned work — you can still use Codex directly for quick one-off tasks or Goal Mode exploration. Chief doesn't lock you out of your own tools.

Does Chief change how Codex runs?

No. Codex still runs in its sandboxed environment, producing changes the same way it always does. Chief provides better context when dispatching and reviews the output when Codex finishes.