AI agents use send_team to create or update resources in Agent Bus — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Agent Bus environment.
This tool creates or modifies data (messages) reversibly in a shared message bus. It does not execute code, delete data, or move financial resources. The blast radius is medium because a malicious agent could spam the team, send misleading instructions to other local agents, or broadcast sensitive information to all team members.
From the tool's definition Tool sends messages to multiple agents ('Send one message to every active agent in a team'), which creates communication artifacts that persist in the message bus. The description explicitly mentions 'fan-out' delivery to multiple recipients.
Attacks that exploit this kind of access
Send one message to every active agent in a team, optionally scoped by project/area. Use for private team fan-out without manually addressing each agent. It is categorised as a Write tool in the Agent Bus MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Agent Bus MCP server in PolicyLayer and add a rule for send_team: 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 Agent Bus. Nothing to install.
send_team 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_team 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_team. 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_team is provided by the Agent Bus MCP server (mustaphasteph/agent-bus). 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 →