AI agents call release_resource to permanently remove resources in Wuying AgentBay — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
Releasing a cloud resource (session, container, or VM) is effectively irreversible: once the environment is torn down, any in-memory state, temporary files, or running processes are permanently destroyed. This maps to Destructive. The blast radius is high because a misfire would prematurely terminate an agent's working environment and lose all unsaved work.
From the tool's definition 'release resource when task finished' — releasing/terminating a cloud execution environment is an irreversible action that destroys the session, any unsaved state, and allocated infrastructure.
Documented attack patterns abuse exactly the kind of access release_resource gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Wuying AgentBay, and nothing reaches the server without passing your rules. This is the rule we recommend for release_resource:
{
"version": "1",
"default": "deny",
"hide": [
"release_resource"
]
}release_resource disappears from the agent's tool list entirely, and any attempt to call it is denied. The rest of the server keeps working.
Free to start. No card required.
release resource when task finished. It is categorised as a Destructive tool in the Wuying AgentBay MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Wuying AgentBay MCP server in PolicyLayer and add a rule for release_resource: 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 Wuying AgentBay. Nothing to install.
release_resource is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the release_resource 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 release_resource. 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.
release_resource is provided by the Wuying AgentBay MCP server (michael98671/agentbay). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Wuying AgentBay, 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.
59 Wuying AgentBay tools catalogued and risk-classified — across an index of 43,000+ MCP servers.