Anthropic adds Claude Code integration to Slack

AIChief

Source

theverge

December 9, 2025

Anthropic is expanding how its Claude assistant works inside Slack by adding direct access to Claude Code. The new feature, now in beta as a research preview, lets Slack users send programming work to Claude Code just by tagging Claude in messages and threads about software issues.

When a user mentions Claude in Slack, the assistant scans the message to identify whether the request involves coding. If it does, Claude automatically routes the task to Claude Code, using context from the Slack thread and any linked code repositories in the web version of Claude Code. This helps developers avoid copying long bug reports, logs, or snippets into separate tools.

Users can also explicitly tell Claude that a message is a coding request. For example, a developer discussing a bug in a project channel can ask Claude to investigate, and Claude Code will analyze the issue using both the conversation and connected repositories. This makes debugging, reviewing code, or understanding older project sections easier without leaving Slack.

The feature works within the existing Claude app for Slack, which previously served mostly as a chatbot for questions and writing assistance. While no new installation is needed, users must still set up Claude Code on the web and connect the repositories they want Claude to access.

The update follows Anthropic’s launch of the Claude Opus 4.5 model. The company claims Opus 4.5 outperforms Gemini 3 on coding tasks, despite the attention Gemini’s benchmarks received. However, safety concerns remain, as early tests showed Opus 4.5 refused only 78 percent of requests related to malware and harmful code creation.