Connects AI agents to Slack workspaces for reading messages, posting updates, and managing channels. Requires careful scope management to prevent data exposure.
Security Rating
81 / 100
Vendor Type
Official (Anthropic)
Risk Category
Medium Risk
Exposed Tools & Function Calls
- slack_list_channels
- slack_post_message
- slack_reply_to_thread
- slack_add_reaction
- slack_get_channel_history
- slack_get_thread_replies
- slack_get_users
- slack_get_user_profile
Required Permission Scopes
- channels:read
- channels:history
- chat:write
- users:read
- groups:read (private channels)
Vulnerability & Risk Analysis
Private channel data exposure (High)
If groups:read and groups:history scopes are granted, agents can read all private channel conversations, exposing sensitive business discussions.
Fix: Grant access only to specific channels using channel-level permissions where possible. Audit all channels an agent can read.
Message impersonation (Medium)
Agents posting as a bot may be mistaken for human employees, enabling social engineering attacks if an agent is compromised.
Fix: Clearly label all bot messages. Add a [AI-Generated] prefix to all agent posts. Never allow agents to impersonate named users.