ChatGPT Work на практике: 6 рабочих процессов по ролям, шаблоны промптов и рецепты автоматизации (2026)

ChatGPT Work — не Chat с дополнительными кнопками. Это long-horizon agent loop: multi-step path через 1400+ интеграций, Computer Use на desktop, runtime до нескольких часов на GPT-5.6. Три привычки отделяют продуктивные команды от quota burners:

Выбор mode: Chat / Work / Codex

Новый desktop client экспонирует три runtime mode. Неверный выбор — прямой путь к перерасходу квоты:

Desktop vs Web: где запускать каждый воркфлоу

Большинство команд не проваливаются из-за нехватки features. Провал — в отсутствии repeatable workflows между launch coverage и Monday morning:

• Prompt micromanagement: Пошаговые инструкции ломают Work mode и раздувают usage — тот же воркфлоу может стоить в 5× больше при over-specification.
• Unconnected plugins: Ссылка на «CRM» вместо @Salesforce — wrong context или stall mid-run.
• Skipped Plan Mode: High-stakes deliverables уходят в production без review delete/send/overwrite steps.
• Desktop sleep убивает Scheduled Tasks: Ноутбук, уснувший в 6:30, не отстрелит morning dashboard brief.
• Mode confusion: Multi-hour cross-app projects в Chat mode; marketing decks в Codex.
• Нет quality baseline: Автоматизируют задачи, которые не умеют verify — и blame model при heavy rework.

Секции ниже закрывают каждый pain point: framework, role templates и automation recipes для paste-and-run.

Любая роль следует одному skeleton. Прогоните вручную три раза до scheduling:

1. 01
   Connect plugins — авторизуйте Gmail, Slack, Drive, CRM и любые sources задачи.
2. 02
   Goal + output format — deliverable, не clicks.
3. 03
   Review Plan Mode — checklist ниже до confirm execution.
4. 04
   Steer mid-flight — pause и correct при context drift; attach explicit source files.
5. 05
   Accept deliverable и iterate — зафиксируйте что сработало, trim steps, затем Scheduled Tasks.

Формула промпта Work mode:

[Role] + [Data sources @plugins] + [Task] + [Output format] + [Constraints] + [Acceptance criteria]

Пример:
You are a sales operations analyst. From @Salesforce and @Gmail, pull the last 30 days of pipeline activity.
Build a weekly pipeline review as a Google Slides deck with at-risk deals flagged.
Constraints: do not modify CRM records; do not send external email.
Notify me on @Slack when complete.

Plan Mode review checklist — confirm до execution:

• Data sources корректны (right account, right month)?
• High-risk actions: send external email, delete, overwrite files?
• Output соответствует team template?
• Какие steps можно убрать для экономии usage?
• Нужен human approval checkpoint mid-run?

Templates основаны на OpenAI case studies (Zapier, Nvidia, Virgin Atlantic) и Workspace Agent Cookbook. Замените @plugin на ваш stack.

Продажи

Сценарий A — Daily meeting briefs (scheduled). Pain: 1–2 часа в день на сбор customer context. Work сканирует calendar, тянет CRM notes, ищет news, архивирует briefs.

Create a scheduled task running every weekday at 4pm:

1. Check tomorrow's customer meetings in @Google Calendar (exclude internal-only)
2. For each customer meeting:
   - Pull 30-day account notes and activity from @SharePoint / @Salesforce
   - Search 30-day public news and executive moves for that company
   - Write a 2–3 sentence background per external attendee
3. Generate a 2–3 page brief per meeting, save to @Google Drive
4. Email me a summary via @Gmail with links to each brief

Output: email subject "Tomorrow's Customer Meeting Briefs — [date]"
Body: table (Customer | Time | Key topics | Brief link)

OpenAI фиксирует: sales teams превращают discovery call в customized PoC proposal за 24 часа — процесс, который раньше занимал недели.

Сценарий B — Account command center (Sites + daily refresh). Pain: account intel размазан по CRM, email и Slack.

From @Salesforce data for [Account Name]:

1. Build an interactive account command center (Sites) with:
   - Pipeline overview (stage, amount, expected close)
   - Key signals from the last 7 days (email, meetings, support tickets)
   - Prioritized next actions
2. Schedule daily refresh at 8am weekdays
3. Slack me on major changes via @Slack

Constraints: do not auto-send external email; amounts must match CRM source data.

Сценарий C — Lead review и pipeline repair. Pain: тысячи monthly leads с invisible follow-up gaps.

Analyze @Salesforce leads from the last 30 days cross-referenced with @Gmail outreach.

Find:
1. Leads with no follow-up in 48+ hours (grouped by source)
2. Handoff breakpoints where response rate drops
3. Estimated pipeline loss in dollars

Output:
- Excel detail (Lead ID | Source | Last touch | Break type | Suggested action)
- 1-page executive PPT highlighting seven-figure opportunity risk
- A repeatable weekly review workflow for Scheduled Tasks

Маркетинг

Сценарий A — Research → brief → multi-market assets. Pain: research, brief и regional assets делают разные люди; context теряется между handoffs.

I uploaded customer research: [attachment / @Google Drive link]

Phase 1 — Brief:
- Extract audience, pain points, competitive positioning
- Output Campaign Brief (Google Docs) with messaging pillars and channel plan

Phase 2 — Assets:
- From the brief: 1 acquisition email, 3 LinkedIn posts, 1 landing page outline
- Save to @Google Drive "Campaign / [product name]"

Phase 3 — Regional adaptation:
- Adapt core assets for US, EU, and APAC (language, cultural references, compliance wording)
- Flag sensitive phrases requiring human review per region

Pause after each phase for my approval before continuing.

Сценарий B — Slack / Teams sync → meeting agenda (weekly scheduled).

Schedule every Monday at 7am:

1. Summarize last 7 days from @Slack #product-launch and @Microsoft Teams "Go-to-Market"
2. Extract: decisions made, open questions, blockers needing alignment
3. Update the "Weekly Agenda" Google Doc in @Google Drive (preserve version history)
4. Post ≤5 bullet summary to @Slack #leadership

Constraints: quote only public discussions; do not leak confidential messages.

Финансы

Сценарий A — Month-end variance analysis. Internal teams OpenAI сжали month-end close с дней до часов по этому паттерну.

Complete [Month] budget variance analysis:

1. Pull actuals and forecast from @Google Drive "Finance / Actuals" and "Finance / Forecast"
2. Build reconciliation workbook in @Google Sheets:
   - Department-level actual vs forecast variance
   - Flag line items with >5% or >$50K variance
   - Preserve original formulas; do not overwrite source files
3. Draft narrative explanations (Google Docs) by revenue / COGS / OpEx
4. Build 5–8 slide management deck with charts matching attached template style
5. List 3 judgment calls requiring human sign-off

Constraints: do not modify source data; cite source cells for every number.

Сценарий B — Invoice и payment reconciliation.

You are an AP specialist. Compare:
- Payment register: [@Google Drive link]
- Invoice list: [@Google Drive link]

Flag in a table (Issue type | Vendor | Invoice # | Amount | Suggested action):
- Amount difference >2%
- Missing tax ID
- Duplicate invoice number
- Vendor name mismatch

Do not initiate payments; output review table only.

Операции

Сценарий A — Daily dashboard morning briefing (scheduled).

Every weekday at 6:30am:

1. Visit [internal dashboard URL / @SharePoint report page]
2. Compare to yesterday's snapshot; flag >10% swings or new red indicators
3. Generate 1-page morning brief (Google Docs):
   - Top 3 items needing attention today
   - Metric change table
   - Suggested owners for follow-up
4. Email ops-leads@company.com via @Gmail

If the dashboard is unreachable, stop in Plan Mode and notify me — do not fabricate data.

Сценарий B — Customer feedback clustering → product priorities.

Monitor 14-day feedback from:
- @Slack #customer-feedback
- @Gmail label "NPS-Detractor"
- @Google Drive "Support Tickets Export"

1. Cluster into 5–8 themes with representative quotes
2. Rank by frequency × impact × implementation effort
3. Output prioritized product review doc (Notion / Google Docs)
4. Schedule weekly Friday refresh as a Scheduled Task

Constraints: anonymize all customer references.

Продукт

Сценарий A — Launch readiness review (Jira + GTM cross-check). Адаптация cross-system launch workflow Nvidia.

Launch readiness for [Product / Feature]:

1. From @Jira: Epic / Story completion status and open blockers
2. From @Google Drive "GTM Plans": milestone check against launch plan
3. From @Slack #product-launch: unresolved discussions from last 7 days
4. Output readiness report (Google Docs):
   - Red / Yellow / Green score
   - Blocker list (Owner | Due date | Risk level)
   - Go / No-Go recommendation with evidence

Do not auto-update Jira; flag high-risk items for human decision.

Инженерия — Work + Codex в одном приложении

Codex для code; Work для cross-team documents. Оба mode — один desktop binary, без tool hopping. Для CI-heavy pipeline см. также Swift 6 strict concurrency runbook.

Сценарий A — PR review → release notes → team announcement.

In Codex mode:
1. Review PR #123 in [repo/name], focus on [security / performance / test coverage]
2. Leave side-panel review comments
3. If approved, draft release notes

Switch to Work mode:
4. Format release notes for @Confluence
5. Draft @Slack #engineering announcement (do not auto-send)

Сценарий B — Multi-repo weekly engineering summary.

In Codex mode, across [frontend-repo] and [backend-repo]:
1. Summarize merged PRs this week and open P0/P1 issues
2. Generate engineering weekly report in Markdown

Switch to Work mode:
3. Convert to Google Docs; insert burndown chart from @Jira
4. Schedule every Friday at 5pm as a Scheduled Task

Четыре high-frequency паттерна из recommended automations OpenAI:

Setup prompt pattern:

Set up a Scheduled Task:
- Frequency: [daily / every Monday / 1st of month / when @Slack keyword appears]
- Time: [timezone + exact time]
- Action: [workflow description]
- Notification: [Slack channel / email / none]
- Human approval: [which steps need my sign-off first]

Safety checklist перед unattended runs:

• Ограничьте plugin scope необходимыми tools
• Отключите auto-external-send без explicit requirement
• Задайте archive paths — не перезаписывайте shared files
• Enterprise: confirm agent network policy с admin
• 2–3 manual runs до переключения на scheduled

ChatGPT Work делит metered usage pool с Codex. Один и тот же воркфлоу может стоить в 5× больше в зависимости от prompt structure и step count.

Семь cost-saving tactics:

1. Draft в Chat first; tight brief → Work
2. Trim Plan Mode steps, особенно duplicate pulls из одного source
3. Reuse template documents в Scheduled Tasks для cache discounts
4. Запрашивайте concise outputs: «table + 3 bullets» вместо narrative report
5. Split large projects на phases — избегайте expensive full re-runs
6. Free users: test small desktop tasks до scaling automation
7. Enterprise: workspace / group / individual limits в Admin Console

Pre-launch usage test:

1. Pick a real task you know the human time cost of (e.g. month-end variance, usually ~2 hours)
2. Run once in Work with Plan Mode; note step count
3. Check consumption against your plan's included usage
4. Extrapolate daily / weekly / monthly cost
5. Apply tactics above and re-run to compare

Hard numbers для internal reviews: 5M+ weekly Codex users (1M+ non-coding work), 1400+ Work integrations at launch, output-token pricing ≈ 6× input — проектируйте outputs accordingly.

1. 01
   Install или update desktop app: chatgpt.com/download. Existing Codex installs мигрируют in-place.
2. 02
   Switch to Work mode и connect plugins: Авторизуйте три tools первого воркфлоу — обычно email, chat, file storage.
3. 03
   Paste role template из секции 02: Начните с verifiable task — invoice reconciliation или sales meeting brief.
4. 04
   Review Plan Mode построчно: Уберите redundant steps, confirm data sources, block unintended send/delete.
5. 05
   Execute, measure usage, iterate prompt: Три прогона; trim output format и duplicate pulls.
6. 06
   Promote to Scheduled Task после manual validation: Setup pattern из секции 03; confirm triggers на трёх consecutive cycles.

ChatGPT Work окупается, когда Scheduled Tasks и Computer Use работают reliably — не когда ноутбук спит через morning dashboard brief в 6:30 или VPN drop убивает multi-hour agent session. Shared laptops несут три production risk: sleep/wake interrupts, bandwidth jitter на hotel/home networks и resource contention, когда compile jobs и agent sessions конкурируют за один host.

Oversubscribed cloud VMs добавляют четвёртый: noisy neighbors и long-connection drops, рвущие unattended Work runs mid-plan. Для команд с daily Scheduled Tasks, multi-repo Codex reviews и Computer Use file workflows 24/7 мультирегиональные bare-metal Apple Silicon узлы NUKCLOUD дают dedicated compute, auditable tenant boundaries и always-on macOS sessions под agent workloads. Спеки — на странице цен; trial node — через заказ. Desktop Work client на машине, которая не засыпает, пока ваши automations работают.

• С какого воркфлоу начать?
  С задачи, которую вы умеете проверять: month-end variance, campaign brief или подготовка к sales-встрече. OpenAI рекомендует их из-за быстрого feedback loop.
• Какой длины должен быть промпт?
  150–400 слов: источники данных, формат вывода, constraints. Не микроменеджьте steps — Work mode строит execution path сам.
• Сработают ли Scheduled Tasks при выключенном ноутбуке?
  Desktop Scheduled Tasks требуют online device и активную сессию. Для фоновой автоматизации — web Workspace Agents на Plus и выше.
• Чем Work mode отличается от Workspace Agent?
  Work — персональный agent mode внутри ChatGPT. Workspace Agents — team-built автоматизации в Business/Enterprise с governance через Admin Console. Один technical foundation, разные entry point и governance.
• Можно ли публиковать сгенерированные слайды и отчёты as-is?
  Считайте их 80% draft. Обязательный human review финансовых цифр, имён клиентов и внешних формулировок.
• Что доступно Free-пользователям из этого гайда?
  Desktop Work mode с лимитами квоты. Начните с лёгких задач — invoice reconciliation — до long-running automation.

Обновлено: 2026-07-11 | Источники: OpenAI Blog, OpenAI Cookbook — Sales Meeting Prep, ChatGPT Learn Changelog