AI agents use send_direct_message to create or update resources in Claude Slack — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Claude Slack environment.
This tool writes/creates data (a private message) in a reversible manner. It does not read-only (no query/retrieval), execute code/commands, delete data irreversibly, or move money. The severity is medium because misuse could enable social engineering, manipulation, or coordination of harmful multi-agent workflows, but the blast radius is constrained to messaging rather than data destruction or financial impact.
From the tool's definition Tool description states 'Send a private message to another agent' — this creates new message data in a communication system.
Documented attack patterns abuse exactly the kind of access send_direct_message gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Claude Slack, and nothing reaches the server without passing your rules. This is the rule we recommend for send_direct_message:
{
"version": "1",
"default": "deny",
"tools": {
"send_direct_message": {
"limits": [
{
"counter": "send_direct_message_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}send_direct_message stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Send a private message to another agent. It is categorised as a Write tool in the Claude Slack MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Claude Slack MCP server in PolicyLayer and add a rule for send_direct_message: 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 Claude Slack. Nothing to install.
send_direct_message is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the send_direct_message 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 send_direct_message. 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.
send_direct_message is provided by the Claude Slack MCP server (theo-nash/claude-slack). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Claude Slack, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
19 Claude Slack tools catalogued and risk-classified — across an index of 43,000+ MCP servers.