Slack Code Enables Collaborative AI Coding
Slack is changing how teams work with AI coding agents by bringing software development into shared conversations. Its new Slack Code feature gives agents a dedicated space where engineers can track progress, provide context, review changes, and collaborate without leaving Slack. This helps avoid situations where agents work in private sessions that leave the rest of the team unable to see the process or contribute important information in time.
When an agent joins a project, the system creates a code channel. Teams can use it to view development plans, code changes, and live previews, while completed channels are automatically archived and remain searchable. Important changes can also be routed to a person for approval directly within the channel, allowing teams to maintain human oversight without giving up automation.
According to Slack, more than seven in ten code channels are completed within a single day. The feature launches with Claude from Anthropic, Devin from Cognition, GitHub Copilot, and Vercel, while ChatGPT from OpenAI is expected to join later. The integrations can, among other things, help teams respond to bugs directly in the conversation and review an agent’s work before it is released to users.
The launch is also part of Slack’s broader move toward collaborative work with AI. The company is introducing an Agents & Tools section for managing interactions with agents across threads, direct messages, and code channels, while also working to make it easier to see when an agent is still handling a task. Slack plans to extend this approach to other business functions in the future.
