Check swarm health status with real state inspection Use when native Task tool is wrong because you need multi-agent coordination — topology (hierarchical/mesh/star), consensus (raft/byzantine/gossip/crdt/quorum), shared memory namespace, or anti-drift gates. For independent one-shot subagents, n...
AI agents call swarm_health to retrieve information from Claude Flow without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and queries the current state of multi-agent swarms (topology, consensus mechanisms, shared memory, drift detection). It performs inspection only and explicitly contrasts itself with Task tools used for actual execution, making it a read-only monitoring/diagnostic function. No data is created, modified, deleted, or executed.
From the tool's definition Tool name 'swarm_health' and description explicitly states 'Check swarm health status' and 'real state inspection' — purely observational operations with no side effects. It inspects topology, consensus, and shared memory state without modifying them.
Attacks that exploit this kind of access
Check swarm health status with real state inspection Use when native Task tool is wrong because you need multi-agent coordination — topology (hierarchical/mesh/star), consensus (raft/byzantine/gossip/crdt/quorum), shared memory namespace, or anti-drift gates. For independent one-shot subagents, native Task is fine; spawn each separately. It is categorised as a Read tool in the Claude Flow MCP Server, which means it retrieves data without modifying state.
Register the Claude Flow MCP server in PolicyLayer and add a rule for swarm_health: 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 Flow. Nothing to install.
swarm_health is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the swarm_health 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 swarm_health. 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.
swarm_health is provided by the Claude Flow MCP server (claude-flow). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.