Notion 3.7 Brings Teamwide AI Skills

Notion 3.7 is one of the clearest examples yet of AI moving from one-off prompting toward repeatable team workflows. Instead of treating prompts as something that only lives inside one person's chat, Notion now lets teams package useful AI behavior into reusable skills.

That is a meaningful shift. In many companies, the best AI workflows are discovered by individuals, but they rarely become part of the team's normal way of working. Notion 3.7 is trying to turn those ad hoc habits into shared operational knowledge.

What Notion 3.7 introduces

The headline feature in Notion 3.7 is teamwide AI skills. Skills are reusable instructions that teach Notion Agent how a team wants work to be done. That could mean formatting a monthly report, reviewing a document with a specific internal standard, drafting a job description using approved language, or transforming notes into a defined output format.

Notion also introduces a shared skills library, downloads for using those skills in external agent environments, more tool connections through Custom MCP, deeper Custom Agent automation, and a redesigned model picker.

AI skills turn prompting into a team asset

The practical value of a skill is not that it saves a single prompt. It is that it captures a process. When a team has a preferred way to review a proposal, prepare an executive brief, or write a recurring document, the skill becomes a reusable standard rather than a hidden trick.

That makes AI more operational. Instead of asking every employee to rediscover the right prompt structure, the organization can define a good workflow once and then reuse it.

The skills library solves a real team problem

Notion's shared skills library is one of the most practical parts of the release. Many useful prompts disappear into chat histories, personal notes, or one employee's memory. A central library gives the workspace a place to store, browse, update, and reuse those skills.

That also creates a versioning advantage. Teams can update a skill in one place instead of having slightly different copies drifting around across several documents and conversations.

Downloadable skills make Notion more portable

One of the more interesting additions is portability. Notion says skills can be downloaded for use in tools such as Claude Code, Codex, Cursor, Gemini, and Grok. Notion writes the instructions as a SKILL.md file together with supporting files.

This matters because teams often do not work inside a single AI environment. A company might standardize skill authoring in Notion while still using different agents and coding tools across departments. A portable skill system makes Notion more useful as a control layer rather than only as an isolated destination.

Custom MCP connections expand what agents can reach

Notion 3.7 also pushes more external context into the agent. The release says users can ask the agent to connect tools such as GitHub and Amplitude, along with other services through Custom MCP connections.

That makes the agent more capable in practice because useful work often depends on information outside the document itself. A workflow that combines notes, analytics, engineering context, and project history is far more useful than one that only summarizes a single page.

Sub-agents and post-meeting automation deepen the system

Notion is also investing in multi-step agent behavior. Custom Agents can now call other Custom Agents as sub-agents, and they can automatically handle work after meeting notes are summarized. That means a larger task can be broken into smaller specialist tasks while still rolling up into one output.

This is an important distinction between an assistant and a workflow system. A chat tool answers a question. A workflow system can hand work off, follow a process, and complete repeated administrative tasks after a trigger.

The release is also about governance and control

Notion 3.7 does not only add power. It also adds more control over how that power is used. Workspace owners can control which models are available, set defaults for Custom Agents, and manage connections and developer tools more clearly.

That kind of control becomes more important as AI shifts from experimentation to team operations. The more these systems are used in real work, the more organizations need visibility into cost, consistency, and which systems are allowed to do what.

Why this matters for professional workflows

The bigger story is not just about Notion. Many tools are moving in the same direction: away from private prompting and toward reusable automation that can be shared across a team. The winners are likely to be the products that turn useful AI behavior into something structured, inspectable, and easy to reuse.

For teams, that means AI becomes more valuable when it behaves less like a novelty and more like part of the operating system of work.

What teams should evaluate before adopting it

Teams should test whether skills are easy to maintain, whether outputs stay consistent, whether permissions are clear, and whether external tool connections are safe enough for the organization's requirements. A shared skill library is powerful only if people can trust it and understand when to use it.

It is also worth checking whether a skill is actually the right level of abstraction. If a process changes every day, a rigid skill may be less useful than a flexible starting point. The most effective skills usually support repeated tasks that already have a recognizable structure.

A step toward reusable AI operations

Notion 3.7 is a strong release because it treats AI as a system of reusable practices, not just a source of answers. If teams want AI to become part of real operations, they need ways to store, improve, share, and govern how it is used. This release moves clearly in that direction.

Sources and further reading

Read Notion's official 3.7 release notes

Explore more creative-tool coverage on the Nejma Blog

Read our Figma Weave article

Visit Nejma

What are Notion AI skills?

They are reusable instructions that teach Notion Agent how to perform a task in a team's preferred way, such as reviewing a proposal, generating a report, or turning notes into a specific output format.

What is the purpose of the skills library?

The skills library gives a workspace one central place to store, browse, share, and update approved AI skills instead of leaving useful prompts scattered across chats and documents.

Can Notion skills be used outside Notion?

Yes. Notion says skills can be downloaded as SKILL.md files with supporting files for use in tools such as Claude Code, Codex, Cursor, Gemini, and Grok.

Why is Notion 3.7 important for teams?

It pushes AI toward reusable team workflows by combining shared skills, external connections, sub-agents, and stronger automation instead of limiting AI to one-off chats.