Slack launches Code channels to let teams build with AI agents in one shared workspace

Slack is rolling out Slack Code, a new way for teams to collaborate with AI coding agents inside dedicated project channels instead of splitting work across chats, docs and external tools. The feature matters because it turns Slack from a communication hub into a more structured place to plan, build, review and approve code with full visibility.

Available starting today on any Slack plan, Slack Code gives teams open, project-specific channels where agents such as Anthropic’s Claude, Cognition’s Devin, Vercel Agent and GitHub Copilot can be invited to help with tasks like shipping a feature, fixing a bug or updating a web page. Slack says the channels are designed to keep the entire process auditable and collaborative, with code diffs, previews and feedback all visible in one place.

The launch is part of a broader race across the tech industry to make AI agents feel less like isolated utilities and more like active members of a team. Slack is clearly betting that developers and product groups want a shared workspace where people and agents can work side by side without losing the history of what was changed, why it was changed and who approved it.

What Slack Code is and why it matters

Slack Code is a dedicated channel format built for AI-assisted software work, and its main value is centralization. Instead of asking a coding agent to generate output in one app, then copying results into Slack for discussion and approval, teams can now keep the entire workflow in one project space.

That workflow is important because AI coding is no longer just about generating snippets. In modern teams, the real challenge is coordination: tracking requests, comparing revisions, reviewing output, sharing context with non-technical stakeholders and preserving a record for later reference. Slack Code tries to solve those pain points by combining conversation, code review and agent activity in a single channel.

Slack says the channel is intended to work like a living project room. A teammate can tag a supported coding agent, the agent can open or join a code channel, and the group can then discuss the task, inspect the result and approve changes before anything ships.

How Slack Code works

Slack Code creates an open space for project-based collaboration with AI agents. When a user needs to build a feature, update content or fix a bug, they can invoke a coding assistant directly in Slack. The agent then generates work inside a dedicated code channel rather than in a separate, disconnected environment.

Within that channel, team members can see the conversation in full, review diffs, preview HTML output and provide comments before giving final approval. Slack is positioning this as a more transparent alternative to private back-and-forth with an assistant that only one person can see.

What happens inside a code channel?

Inside a code channel, the work is meant to be visible from start to finish. Slack says users will be able to audit code changes, inspect live previews of what the agent produced and weigh in before the output is merged or published.

The company is also emphasizing accountability. Because the channel is tied to a specific task, it can serve as both a collaboration space and a record of how the work was completed.

• Task kickoff: A user tags a supported coding agent with a request.
• Channel creation: The agent opens or uses a dedicated project code channel.
• Visible collaboration: Team members review discussion and progress in real time.
• Diff review: Changes can be compared before approval.
• Preview step: HTML output can be checked before shipping.
• Archive: The channel closes automatically once the assignment is complete.

Who can use Slack Code?

Slack Code is available on any Slack plan, which means the feature is not being limited to a premium enterprise tier at launch. That broad availability suggests Slack wants rapid adoption among small teams, startup developers and larger organizations that already use Slack as their day-to-day operating layer.

The feature works with AI agents available through Slack’s marketplace. Slack says founding partners include Claude Code, Devin, Vercel Agent and GitHub Copilot, and that those tools will integrate smoothly with code channels.

This partner list is significant because it shows Slack is not trying to replace coding tools already in use. Instead, it is trying to become the collaboration surface where those tools are orchestrated.

Supported agents at launch

Slack’s initial lineup indicates a strategy built around interoperability rather than a proprietary agent. That may help the company appeal to organizations that already have preferences for certain coding systems or want to compare assistants across tasks.

Why Slack is betting on AI agents as teammates

Slack’s message is not just about productivity; it is about changing how teams think about AI. The company is framing coding agents as collaborators that can take on work in a shared project space, almost like another teammate in the channel.

That framing lines up with a growing trend in software: companies are increasingly presenting AI systems as semi-autonomous helpers rather than passive chat tools. Slack’s approach gives that trend an organizational structure. The goal is to make the agent’s work visible, reviewable and accountable, which is especially important in software development where small errors can become costly quickly.

Slack also appears to be responding to a wider market shift in which users expect AI tools to do more than answer prompts. Teams want agents that can act, remember context and fit into existing processes. By building around channels, Slack is using one of its core product strengths—persistent shared conversation—to support that model.

How this compares with other agent strategies

Slack is not alone in pushing the “AI colleague” idea. Other companies are also trying to make agents feel embedded in real workflows rather than isolated chatbots.

The difference is that Slack has a uniquely collaborative setting already in place. If a team already uses Slack to coordinate work, a code channel can become the default home for agent-driven projects without requiring people to adopt a new platform first.

Slack’s pitch is that teams should be able to invite an AI coding assistant into the same conversation where the work is planned, reviewed and approved, instead of forcing everyone to juggle separate tools and fragmented records.

What makes Slack Code different from ordinary chat-based coding help?

Slack Code is different because it is designed around a project lifecycle, not a one-off prompt. Standard AI coding help often begins and ends in a private interaction. Slack Code instead structures the interaction in a shared channel with visible steps, making the process easier to manage for teams that need collaboration and oversight.

That matters for both technical and non-technical users. Developers can evaluate the code, while designers, product managers and operations staff can follow the discussion, understand what changed and contribute feedback before release.

The automatic archiving feature also suggests Slack wants these channels to be temporary workspaces tied to specific jobs. Once the assignment is done, the channel closes itself and leaves behind a record that can be searched later if needed.

What does the feature mean for software teams?

For software teams, Slack Code could reduce the friction between idea, implementation and review. Many teams already coordinate tasks in Slack, then move to development tools for execution and back again for approvals. Slack Code tries to pull those stages together.

The immediate benefit is clearer visibility. A bug fix or feature request can be traced from the initial message through the agent’s output and final sign-off. That may reduce the chance of losing context across tools or missing why a particular code path was chosen.

There is also a governance angle. Audit logs and review steps could make Slack Code attractive to teams that need stronger documentation around AI-assisted work. As more companies adopt agents, leadership teams are likely to care not just about speed but also about traceability.

Potential advantages for teams

• Less context switching between chat, review and coding tools
• Better transparency into what the AI changed
• Easier collaboration among technical and non-technical staff
• Improved recordkeeping for approvals and revisions
• Faster iteration on routine coding tasks

Potential limits and questions

Slack’s announcement leaves some practical questions unanswered, including how deeply the code channels will integrate with existing developer workflows, how permissions will work across larger organizations and how much control administrators will have over agent behavior.

Those details may matter a great deal if Slack Code becomes a serious part of enterprise engineering operations. Teams will likely want to know how outputs are validated, whether channels can connect to repositories and how secure the full workflow is when multiple agents and users are involved.

Why the launch is strategically important for Slack

Slack has long been a workplace communication tool, but the company is under pressure to stay central as AI changes how knowledge work gets done. By adding Code channels, Slack is trying to ensure that its platform remains the place where important work happens, not merely the place where people talk about it.

The move also helps Slack position itself against the growing number of AI-native tools that want to own project collaboration from start to finish. If agents become a standard part of team work, the platforms that host the conversation around them could become just as valuable as the agents themselves.

In practical terms, Slack is betting that teams will prefer one visible space where they can coordinate humans and AI together. That could strengthen the company’s role in the software development workflow, a category that is becoming increasingly crowded and increasingly AI-driven.

Timeline of Slack Code’s launch

The rollout is straightforward, but the sequence helps show how Slack is framing the feature as immediately usable rather than experimental.

How Slack Code fits the broader AI agent trend

Slack Code reflects a wider industry belief that the next phase of AI will be agentic, collaborative and workflow-aware. Instead of simply responding to questions, AI systems are increasingly being asked to help complete tasks, coordinate with people and operate inside business tools.

That shift is visible across software, design and operations products, where vendors are racing to make AI feel like an embedded participant in everyday work. Slack’s contribution is to bring that idea into a familiar communication environment with an emphasis on transparency and teamwork.

If the feature catches on, it could encourage more companies to treat AI work as a documented, shared process rather than an invisible exchange between one employee and a model. That would be a meaningful change for how teams manage software creation and review.

What to watch next

The most important question now is whether teams actually adopt code channels as a routine part of development. Slack has an advantage because it already sits at the center of many workplace conversations, but the success of Slack Code will depend on how well it integrates into day-to-day engineering habits.

Watch for whether companies use the feature for small fixes, content updates and experimental builds first, then expand into larger development work. Also watch how strongly Slack pushes the audit and approval angle, since that may become a selling point for compliance-minded businesses.

For now, the company has made a clear bet: if AI agents are going to become teammates, they need a shared room to work in. Slack wants that room to be Slack.