saved_delete
AI agents call saved_delete to permanently remove resources in Slack — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
The tool name strongly suggests permanent deletion of saved Slack content. Although the description is empty, the semantic meaning of 'delete' in a data context indicates an irreversible destructive action. Given the server's scope (Slack workspace access with 220 tools), misuse could result in loss of important saved messages or organizational knowledge.
From the tool's definition Tool name 'saved_delete' indicates deletion of saved items in Slack. The 'delete' verb combined with 'saved' (referring to saved messages, searches, or bookmarks) suggests irreversible removal of user data.
Attacks that exploit this kind of access
saved_delete. It is categorised as a Destructive tool in the Slack MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Slack MCP server in PolicyLayer and add a rule for saved_delete: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Slack. Nothing to install.
saved_delete is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the saved_delete rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for saved_delete. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
saved_delete is provided by the Slack MCP server (karbassi/slack-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →