Slack users can now connect with Anthropic’s Claude Code directly within the platform by tagging Claude on coding-related messages and threads. This feature is being introduced in beta as part of a research preview.
When a user tags Claude, the system will automatically analyze the message to determine if it pertains to a coding task. If so, it will forward the request to Claude Code while using context from the associated Slack thread and authenticated code repositories. Users can also specify directly that their request is a coding task, such as asking Claude to examine a bug report without needing to manually transfer all relevant information into Claude Code.
This new functionality is an extension of the existing Claude app for Slack, which previously operated as an in-app chatbot. No additional downloads are necessary for those who already have the app, although it is required to set up the web version of Claude Code with desired code repositories.
The rollout of this feature occurs shortly after Anthropic introduced its Claude Opus 4.5 model. The company asserts that the latest version of Claude excels in coding compared to Google Gemini 3, whose initial benchmarks prompted significant industry attention. However, concerns remain regarding the safety and security of the model; early testing indicated that Opus 4.5 only denied 78 percent of requests to generate malware and other harmful code.
This integration raises questions about the potential implications for coding practices and security within collaborative platforms like Slack.
Source: https://www.theverge.com/news/839817/anthropic-claude-code-slack-integration