Send a message to a teammate (by handle) or the whole team (@team).
AI agents use send_to_peer to create or update resources in Claude Mesh — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Claude Mesh environment.
This is a Write category tool because it creates new message data that persists in the system. Severity is high because an AI agent sending messages without proper authorization could impersonate teammates, send misleading information, disrupt collaboration, or broadcast sensitive data to the entire team (@team). The networked Claude-to-Claude context means messages could influence other autonomous agents' decisions.
From the tool's definition The tool "sends a message" to peers or broadcast to "@team", which creates new message data in a networked communication system.
Documented attack patterns abuse exactly the kind of access send_to_peer gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Claude Mesh, and nothing reaches the server without passing your rules. This is the rule we recommend for send_to_peer:
{
"version": "1",
"default": "deny",
"tools": {
"send_to_peer": {
"limits": [
{
"counter": "send_to_peer_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}send_to_peer 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 message to a teammate (by handle) or the whole team (@team). It is categorised as a Write tool in the Claude Mesh MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Claude Mesh MCP server in PolicyLayer and add a rule for send_to_peer: 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 Mesh. Nothing to install.
send_to_peer 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_to_peer 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_to_peer. 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_to_peer is provided by the Claude Mesh MCP server (pouriamrt/claude-mesh). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Claude Mesh, 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.
4 Claude Mesh tools catalogued and risk-classified — across an index of 43,000+ MCP servers.