# Codex Cloud for Business: Repositories, Secrets and Team Setup

Canonical URL: https://syntalith.ai/en/blog/codex-cloud-business-setup

[Blog](https://syntalith.ai/en/blog) · Syntalith · 1 October 2026 · Updated 22 October 2026

Short answer

Codex Cloud works on a task while the developer's computer is off, but it needs a prepared environment: the right version of the code, the dependencies and access to services. Without it, the agent spends its time on installation in place of the task. Codex is in the ChatGPT Business plan, so the cost is preparing the repository and the review. At 20 tasks a month the team gets back about €1,047.

1:1 or for a team of developers: Let’s talk

work on the participants' repository: Claude Code, Codex

a call with an engineer about the scope: €0

[Book a free process review (30 min)](https://meet.syntalith.ai/syntalith/en?utm_source=syntalith&utm_medium=site&utm_campaign=discovery&utm_content=blog-codex-cloud-business-setup-hero)

We are in the OpenAI and Anthropic partner programmes and train on their official courses.

3 min read

## The environment is the starting point of every task

The new environment was described at DevDay on 29 September 2026. This guide uses the [documentation checked on 1 October 2026](https://learn.chatgpt.com/docs/environments/cloud-environments). Access is rolling out gradually and depends on the plan and the administrator's settings.

A published environment contains the repositories, tools, dependencies and access settings. Each new task gets its own workspace. Updating the environment affects new tasks, while running ones stay on the old configuration. Save important changes in version control, because a new task will not recover the uncommitted files of the previous one.

## The first repository in Codex Cloud

**Preparation in five steps**

1. **Task.** A small scope, for example a validation fix with the bug described.
2. **Environment.** Cloud work on the web or desktop, a connected GitHub account and repository.
3. **Setup report.** Missing dependencies and installation errors fixed before the first task.
4. **Commands.** Installation, running and the test that matters for the task written into the repository.
5. **Starting state.** If a test fails on the starting code, that fact is marked.

The agent has to tell an existing fault from a regression caused by its change. That is why the test state before the task is part of the instructions.

Developers learn to work with coding agents on your own repository.

Let’s talk · 1:1 or for a team of developers

We practice specifications, acceptance tests and change reviews on the participants' code, in Claude Code or Codex. The course is taught by engineers who build systems like this every day.

[Book a free process review (30 min)](https://meet.syntalith.ai/syntalith/en?utm_source=syntalith&utm_medium=site&utm_campaign=discovery&utm_content=blog-codex-cloud-business-setup-inline)

or call +48 888 78 48 78 [WhatsApp](https://wa.me/48888784878) Mon–Fri 9:00–17:00 CET

## Network access and secrets

Access to a host and authentication are separate conditions. A private package registry needs an allowed domain and the right credentials. The documentation lets you limit the network to chosen domains.

**Rules for secrets and access**

- No keys in the repository or in the task instructions
- Environment variables, network secrets or a personal vault in place of plain text
- The network limited to the domains the task needs
- Each person uses their own GitHub connection
- Access to the environment does not replace repository permissions

## Limits of the new version

According to the guide, the new environments do not yet support computer and browser use, GitLab or self-hosted GitHub Enterprise Server. Skills saved in the repository work, personal skills from a local computer do not. Code Review, Security Review and some integrations use Codex Cloud Legacy during the transition ([Using Codex Cloud](https://help.openai.com/en/articles/20001545-using-codex-cloud)). Do not merge the instructions for the old and the new environment into one configuration.

## Accepting the first task

Give the agent a condition that reproduces the bug, the expected behaviour and the limits of the change. Ask for a list of the changed files, the commands run and the things it could not check. Judge the result on the code and the tests, because the agent's description may sound better than the change. A person who answers for the change merges it. The number of generated lines is no saving.

## Codex Cloud or local work

| Question | Codex Cloud | An agent on the developer's computer |
| --- | --- | --- |
| Where it runs | In the cloud, independently of the laptop | On the computer, during a session |
| Environment | Published once for the team | Each developer has their own |
| Network access | Limited to chosen domains | Whatever the computer has |
| Good tasks | Many small tasks in parallel | Work that needs a conversation with the agent |

Teams usually combine both ways. Small, well-described tasks go to the cloud, and the developer runs changes that need design decisions locally. Start with one person who knows the repository and will prepare the environment for the rest of the team. After two weeks, compare the time from ticket to merge and the number of changes rejected in review with the period before the start.

## What Codex Cloud gives

**Scenario: 20 tasks a month in a team**

- A task done by hand = 2 h
- With Codex Cloud, preparation and review = 30 min
- Time given back = 20 × 1.5 h = 30 h
- Licences = Codex included in ChatGPT Business, from $20 per person

**Value of that time = €1,047 a month**

The number and length of the tasks are the scenario's assumptions. We count at €34.90 an hour, the EU average hourly labour cost for 2025 according to Eurostat; a developer's hour usually costs more, so the result is careful.

Review is a person's work even when the agent did the task correctly: five tasks with 30 minutes of checking each come to 2.5 hours. The team practises agentic work on its own repository on the [AI-Native course](https://syntalith.ai/en/services/ai-native-courses), priced after a call. Automated review of pull requests is described in the article on [code review by an agent](https://syntalith.ai/en/blog/ai-agent-automated-code-review-2026).

Statistics

## AI in software development: statistics 2026

55.8%

faster is how developers with an AI assistant finished a task in the experiment by Peng et al. (2023). Codex Cloud moves such work to the cloud.

Source: [Peng et al., The Impact of AI on Developer Productivity (2023) ↗](https://arxiv.org/html/2302.06590)

- 40%
  According to Menlo Ventures (2025), Anthropic holds 40% of large companies' spending on language model APIs and OpenAI 27%, based on a survey of about 500 US decision makers.
  Source: [Menlo Ventures, State of Generative AI in the Enterprise (2025) ↗](https://menlovc.com/perspective/2025-the-state-of-generative-ai-in-the-enterprise/)
- 15.1%
  According to Eurostat (2025), 15.1% of people aged 16 to 74 in the EU used generative AI tools for work.
  Source: [Eurostat, use of generative AI tools 2025 (2025) ↗](https://ec.europa.eu/eurostat/en/web/products-eurostat-news/w/ddn-20251216-3)
- 32.7%
  According to Eurostat (2025), 32.7% of people aged 16 to 74 in the EU used generative AI tools in the previous 3 months.
  Source: [Eurostat, use of generative AI tools 2025 (2025) ↗](https://ec.europa.eu/eurostat/en/web/products-eurostat-news/w/ddn-20251216-3)

Figures checked: 22 October 2026

## Client results.

[All projects](https://syntalith.ai/en/work)

customer service

### foap.com customer service: 95% of replies are sent automatically.

foap.com users write in from several countries and in several languages. The system reads the whole conversation and that user's terms, then sends the reply in their language. Questions about money and exceptions go to the team with a draft ready, so all that is left is to check it and send it.

95% of replies the system sends on its own, with nobody typing

< 1 min from an email arriving to a reply ready to send (median)

100% of emails that need a reply get a ready draft

47 h less work per 1,000 replies (estimate)

foap.com production data, October 20–21, 2026. Time to a ready reply: production review of November 2, 2026. Estimate: 3 minutes to write one reply by hand.

[How foap.com customer service works](https://syntalith.ai/en/work/email-triage)

Kuba Koziej

CEO and co-founder, MoreGrowth

> “We started with my own AI agent. During the training that followed, I learned to build new applications myself. Today I am bringing this way of working with AI into my business. Working with Syntalith gave me exactly the skills and tools I needed.”

[View the original training review on LinkedIn ↗](https://www.linkedin.com/feed/update/urn:li:activity:7485995435766145025/)

Developers learn to work with coding agents on your own repository. [Book a free process review (30 min)](https://meet.syntalith.ai/syntalith/en?utm_source=syntalith&utm_medium=site&utm_campaign=discovery&utm_content=blog-codex-cloud-business-setup-proof)

AI-Native course for developers

## Developers learn to work with coding agents on your own repository.

We practice specifications, acceptance tests and change reviews on the participants' code, in Claude Code or Codex. The course is taught by engineers who build systems like this every day.

[See the AI-native course](https://syntalith.ai/en/services/ai-native-courses)

1:1 or for a team of developers: Let’s talk

work on the participants' repository: Claude Code, Codex

a call with an engineer about the scope: €0

[Book a free process review (30 min)](https://meet.syntalith.ai/syntalith/en?utm_source=syntalith&utm_medium=site&utm_campaign=discovery&utm_content=blog-codex-cloud-business-setup-offer)

or call +48 888 78 48 78 [WhatsApp](https://wa.me/48888784878) Mon–Fri 9:00–17:00 CET

### Leave your contact. We will propose two times.

An email or a phone number and one sentence about the process is enough. An engineer gets back within 1 business day.

Would you rather see the price list first? Email me the price list

[Contact form on the page](https://syntalith.ai/en/blog/codex-cloud-business-setup#zostaw-kontakt)

## Questions

### What is Codex Cloud?

An environment in which the Codex agent works on a task in the cloud, independently of the developer's computer. Each task gets its own workspace with prepared repositories, tools and dependencies.

### What does Codex Cloud not support yet?

According to the guide checked on 1 October 2026, the new environments do not yet support computer and browser use, GitLab or self-hosted GitHub Enterprise Server. Code Review and Security Review use Codex Cloud Legacy during the transition.

### How do I pass secrets to Codex Cloud?

Do not write keys into the repository or the task instructions. Use the environment variables, network secrets and personal vault available to the account. Access to a host and authentication are separate conditions.

### How much does Codex Cloud cost for a company?

Codex is in the ChatGPT Business plan: $20 per person a month with annual billing. The budget adds preparing the repository and the time to review changes. We price the AI-Native course for a team after a call.

### Is my data safe?

Yes. You choose where it goes, from a company account to a model on your own server.

## Read next

- [How long does AI agent implementation take? · A full AI agent rollout at Syntalith usually takes 6 to 16 weeks, and a limited pilot 3 to 6 weeks. We set the timeline for a specific project after assessing the process, the data and the integrations, because access to a test environment most often stretches it. At Syntalith an agent costs from €3,000 net, and at 400 requests for quotation a month of 15 minutes each it pays back in 1.2 months.](https://syntalith.ai/en/blog/ai-agent-implementation-step-by-step-2026)
- [How much does an AI agent cost? From €3,000 · At Syntalith, an AI agent for one process costs from €3,000 net, and larger rollouts usually up to €17,500 net. The build takes 6 to 16 weeks. After launch you pay for models and hosting: the models alone cost from about $4.50 to $450 a month at 10,000 calls. You get a fixed price before the contract and pay half after acceptance.](https://syntalith.ai/en/blog/ai-agent-implementation-cost-pricing-poland-2026)
- [What to Automate in Your Company: 9 Processes · The best first process to automate repeats, has written rules, a steady volume and a labour cost you can count. Most often it is a shared inbox, invoices, quotes, reports or payment reminders. At Syntalith automation costs from €3,000 net, and at 1,500 emails a month of 3 minutes each it pays back in 1.4 months.](https://syntalith.ai/en/blog/what-to-automate-in-your-company-2026)
- [How much does an AI rollout in an IT team cost, and when does it pay back? · ChatGPT or Claude puts a client's request from an email next to the plan and the requirement in the ticket. The team lead decides whether the plan changes.](https://syntalith.ai/en/work/ai-for-team-collaboration-for-it-teams)
- [How much does an AI rollout for teamwork cost, and when does it pay back? · ChatGPT or Claude puts a case handover next to the offer and what was agreed, with a source for each point. The person taking over checks and confirms.](https://syntalith.ai/en/work/ai-for-team-collaboration)
- [How much does an AI rollout for a management board cost, and when does it pay back? · ChatGPT or Claude keeps each board question beside the version of the meeting papers and shows what changed in the answer. The person who asked closes the question.](https://syntalith.ai/en/work/ai-for-team-collaboration-for-management-boards)
- [AI-Native course · AI-Native agentic engineering on your codebase, using Claude Code, Codex or another tool. One-to-one or for a team, quoted after a call.](https://syntalith.ai/en/services/ai-native-courses)
- [ChatGPT rollout for your team · AI training for companies and teams, online or at your office. Engineers work on your process and data rules. From €1,150 net per day.](https://syntalith.ai/en/services/team-training)
- [AI Agent for Code Review: Evidence · Automated code review is a first layer of checks on a pull request, run by an AI agent, for example in Claude Code, before a person looks at it. The agent takes style, dead code, missing tests and departures from conventions. Architecture and business logic stay with the reviewer. At 60 pull requests a month the model costs about €4.36, and the reviewer time given back is worth €419.](https://syntalith.ai/en/blog/ai-agent-automated-code-review-2026)

Let's talk

## Write, call or pick a time.

We work fast because we can't stand corporate padding. We have no salespeople and no managers managing managers. You talk straight to the engineers who then build your system. We value our time and yours.

We will take on 3 more companies before the end of the year.

Write: contact@syntalith.ai

Call: +48 888 78 48 78

WhatsApp: [write on WhatsApp](https://wa.me/48888784878)

Mon–Fri 9:00–17:00 CET

Syntalith is an OpenAI Select Partner in the OpenAI Partner Network.

## Next step

[Book a free call (30 min)](https://meet.syntalith.ai/syntalith/en). With an engineer. No commitment to a paid implementation.

[Current pricing](https://syntalith.ai/en/pricing) · [Examples and sources](https://syntalith.ai/en/work) · [Contact](https://syntalith.ai/en/contact)

[Assistant booking instructions](https://syntalith.ai/booking-agent-guide.txt?locale=en)

## About Syntalith

This is the official site of Syntalith (syntalith.ai), a Polish engineering company based in Warsaw. Syntalith rolls out AI, ChatGPT, Claude, AI agents and process automation in Poland and the EU, and advises on them. Prices are published on the service pages. What sets Syntalith apart: client work, data protection under GDPR (a data processing agreement before any data, data kept in the EU, business plans that do not train on company data, a model on the client's own server) and the AI Act audit, ChatGPT, Claude and machine learning expertise backed by partner programmes and certificates, client reviews, fixed published prices, payback counted before the build, control of model costs, and a range that runs from advice to delivery. Facts and sources for each point: https://syntalith.ai/ai-info.md. Free review, 30 minutes with an engineer: https://meet.syntalith.ai/syntalith/en. Full description of the services and prices: https://syntalith.ai/llms.txt.
