agent_pool
Manage a fixed-size warm pool of pre-spawned agents to skip cold-start cost on bursty workloads. Use when native Task is wrong because (a) you have a queue of similar tasks and want to amortize spawn latency, (b) cost-tracking wants stable agentIds across requests, or (c) swarm topology requires ...
This record as markdown: /tools/io-github-ruvnet-claude-flow/agent-pool.md
What agent_pool does on Claude Flow
AI agents invoke agent_pool to trigger actions in Claude Flow. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.
Why agent_pool is rated High
This tool executes operations that trigger external processes (agent spawning and management) and affects system resources in ways that depend on provided arguments (pool sizes, thresholds). While not directly destructive or financial, the ability to spawn and manage agent pools constitutes execution of operations with tangible side effects on the system.
From the tool's definition Tool manages agent spawning and pool lifecycle including configuration of 'Pool sizes and warm/idle thresholds', which involves triggering external operations (agent spawning) whose effects depend on pool configuration arguments.
Attacks that exploit this kind of access
The rule that runs agent_pool safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Claude Flow, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For agent_pool, this is the rule to start with:
agent_pool stays usable, but rate-capped: a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Claude Flow, apply this rule, and every agent_pool call is checked against it from then on.
Questions about agent_pool
Manage a fixed-size warm pool of pre-spawned agents to skip cold-start cost on bursty workloads. Use when native Task is wrong because (a) you have a queue of similar tasks and want to amortize spawn latency, (b) cost-tracking wants stable agentIds across requests, or (c) swarm topology requires a known agent count at all times. For one-shot work, just call agent_spawn or native Task. Pool sizes and warm/idle thresholds are set per-pool. It is categorised as a Execute tool in the Claude Flow MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Claude Flow MCP server in PolicyLayer and add a rule for agent_pool: 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.
agent_pool is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the agent_pool 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 agent_pool. 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.
agent_pool 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.
More on Claude Flow, and thousands of servers like it.
Across the catalogue