Назад към Новини
ПродуктAI Understanding брифинг

Google пуска Gemini умения за автоматизиране на повтарящи се подкани

Google обяви, че чатът Gemini вече ще поддържа „Умения“, функция, която позволява на потребителите да запазват и използват повторно персонализирани инструкции, заменяйки по-стария инструмент „Скъпоценни камъни“.

4 min readRead the primary source
Source-provided image accompanying Google rolls out Gemini Skills to automate repetitive prompts
Документ с първичен източникИзточникът е записан
Издател
blog.google
Изходна връзка
blog.googlehttps://blog.google/products-and-platforms/products/gemini/automate-tasks-with-skills/
Тип източник
Първичен документ — официално съобщение, документ, документ или първа страна, която четем директно.
КонтекстРазберете това за 60 секунди

Започнете тук

Ключови термини

API (интерфейс за програмиране на приложения)
Структуриран начин за една софтуерна система да изпраща заявки до и да получава отговори от друга система.
Prompt Engineering
Проектиране на подкани за подобряване на качеството на изхода, надеждността и контролируемостта.
Характеристика
Входна променлива, използвана от модел за правене на прогнози.
Тествайте себе сиКакво е AI? Тест

Какво стана

Google is launching a new called Skills in Gemini chat. Skills let users save frequently used prompts as reusable commands that can be invoked with a forward slash (/) and the skill name. The feature is live globally for Gemini Spark users and will be extended to Google Workspace customers in the coming weeks. Skills can combine multiple instructions, include reference files such as PDFs or images, and will inherit capabilities previously offered by Gems, such as sharing and Drive integration. Google will automatically migrate existing Gems into Skills and will discontinue Gems for personal accounts in November, for Workspace business, enterprise, and nonprofit customers in March 2027, and for education customers in June 2027. The related Opal 2 mini‑app experiment will also be retired at the same time.

Google announced that Skills are now available in Gemini Spark and will be rolled out to Google Workspace customers in the coming weeks. Users can create a skill by saving a prompt and later invoke it with a simple '/' command.

Skills can be combined, allowing users to apply multiple instructions—such as a writing style and brand guidelines—simultaneously. They also support attaching reference files like plain‑text documents, PDFs, or images, extending Gemini’s ability to work with external content.

Google will automatically migrate existing Gems into Skills, and will discontinue support for Gems starting November for personal accounts, March 2027 for business, enterprise, and nonprofit accounts, and June 2027 for education accounts. The Opal 2 mini‑app experiment will be retired at the same time.

Детайли за източника: blog.google ↗

Защо има значение

The introduction of Skills turns Gemini chat into a more practical productivity tool by reducing the need for repetitive prompting, which can speed up routine content creation, data extraction, and brand‑compliant writing. By allowing reference files and the stacking of multiple skills, the expands Gemini’s utility beyond simple chat, enabling more complex, repeatable workflows that can be shared across teams. This shift also signals Google’s broader strategy to embed AI capabilities directly into its Workspace suite, potentially increasing adoption among enterprise and education users. However, the phased retirement of Gems and the Opal 2 experiment means existing users must transition their custom automations, and the exact pricing or licensing terms for Skills in Workspace environments have not been disclosed.

By reducing repetitive prompting, Skills can improve productivity for individuals and teams that rely on Gemini for drafting, summarizing, or generating on‑brand content.

The ability to attach reference files and stack skills expands Gemini’s functional scope, making it more suitable for complex tasks that previously required manual .

The phased retirement of Gems forces users to adopt the new workflow, which may drive higher engagement with Gemini’s AI capabilities across Google’s enterprise ecosystem.

Interactive Mechanism

Интерактивен механизъм: как всъщност работи

Разгледайте интерактивно основната технология зад тази разработка.

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:🛡️ Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
Интерактивна проверка на концепцията+10 Points
What is AI? Quiz

A route planner searches possible journeys using explicit rules. What does this illustrate about AI?

Какво да гледате след това

Future updates on the pricing model for Skills in Workspace, the timeline for full parity with Gems, and any performance metrics Google releases about user adoption. Watch for developer documentation that may enable third‑party integrations or custom skill marketplaces, as well as any feedback from early adopters about workflow efficiency gains or limitations.

Pricing and licensing details for Skills within Google Workspace, especially for enterprise and education customers.

Potential release of an API or marketplace that would let developers publish and monetize custom Skills.

User adoption metrics and feedback that could indicate whether Skills achieve the intended efficiency gains.

Any further integration of Skills with other Google AI products, such as Search or Docs, which would broaden the impact of the .

Свързани ръководства и викторини

Какво е AI?Обяснени модели на AIPrompt EngineeringБъдещето на ИИТествайте какво знаете — опитайте безплатен тест с изкуствен интелектПотърсете термин за AI в нашия речникСледвайте програмата за проследяване на пускането на AI модел
Намирате ли това за полезно?