Slack has introduced Slack Code, a dedicated workspace for teams to collaborate with AI coding agents on software development. Code channels bring coding work into Slack, allowing developers and other team members to share context, review changes and work with agents in a common environment.
Coding agents such as Claude Code, Devin, GitHub Copilot, ChatGPT and Vercel agents have made software development faster and more accessible. However, much of the work between an individual and an AI agent currently happens outside team conversations, which can result in lost context, delayed reviews and additional hand-offs.
Slack Code is designed to bring that work into a dedicated space within Slack. Each code channel is associated with a specific project, allowing teams to work with an agent on coding tasks while keeping the conversation, code and related context together.
Code channels can be created automatically when a coding agent is tagged from an existing conversation. Once created, the channel provides a shared view of the agent’s work, including the conversation, plan, code changes and live previews. Completed channels can be archived while retaining their history as an audit record.
Teams can use Slack Code when developing a new feature, updating a webpage or fixing a bug. A user can bring a coding agent into the task, after which team members can follow the work, review code differences, inspect live previews, provide feedback and approve changes before they are shipped.
The shared workspace is also intended to bring non-technical team members into software development. Product managers, designers and other employees can participate in the process without opening a terminal, while engineers can review and approve the resulting code.
For example, a product manager can identify a bug in a Slack channel and ask a coding agent to help develop a fix. The agent can use the existing conversation, screenshots and documents as context, propose a change and create a code channel. An engineer can then review the code, run the preview and instruct the agent to open a pull request or merge the change.
Slack Code is being developed with an ecosystem of coding-agent providers. Anthropic, Cognition, GitHub, OpenAI and Vercel are among the companies developing integrations, allowing their agents to work within the shared Slack environment. Slack also plans to open its code-channel APIs to the broader developer community, enabling custom agents to participate in the workspace.
Slack is also expanding how agents appear across its platform. Agent DMs provide dedicated conversations with clearly identified agents, while a new Agents tab gives users a central place to view agents and their sessions, check statuses and stop agents when required. The Add to Slack capability enables users to bring agents built with platforms including Lovable, Vercel, OpenAI, LangChain and n8n into Slack.
Security and administrative controls are built into the agent experience. Agents in code channels inherit Slack’s existing security model, permissions and administrative controls. For high-stakes actions such as pushing code to production, the agent packages its work for human review and sign-off within the channel.
“AI only creates value when it’s part of how a team actually works, not something people go do alone in another tab. Code channels bring that work into the open, where the rest of the team can see it, shape it, and build on it,” said Rob Seaman, EVP and GM of Slack, Salesforce.
Slack said the initial release connects with Anthropic’s Claude, Cognition’s Devin, GitHub’s Copilot, OpenAI’s ChatGPT and Vercel’s agents. Slack Code is available across Slack plans, with access to the respective partner agent required.
The company said Slack Code represents a shift towards collaborative software development in which humans and AI agents work within the same shared environment, retaining the context and history of the development process.