call_action · AgentBridge MCP Server MCP: Risk & Policy

WHAT IT DOES

What call_action does on AgentBridge MCP Server

AI agents invoke call_action to trigger actions in AgentBridge MCP Server. 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 IT NEEDS A POLICY

Why call_action needs a policy

This is a dynamic action invocation tool that can trigger arbitrary actions from a manifest. While sibling tools include Financial (refund) and Destructive (implied delete operations) options, 'call_action' itself is the generic executor.

From the tool's definition Tool description states it can 'Invoke an AgentBridge action' which triggers external operations. The confirmation mechanism and idempotencyKey pattern indicate this tool executes actions with real side effects whose outcomes depend on which action is invoked…

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

call_action Rate-limited

call_action 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.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "call_action": {
      "limits": [
        {
          "counter": "call_action_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

RATE-LIMIT THIS TOOL →

Instant setup, no code required.

PolicyLayer is an MCP gateway: every tool call is checked against your rule before it reaches AgentBridge MCP Server, so you can run agents against it safely.

FAQ

Questions about call_action

What does the call_action tool do? +

Invoke an AgentBridge action. Risky actions return a confirmationRequired response with a confirmationToken; the client must re-call with confirmationApproved: true AND the same confirmationToken to execute. Optional idempotencyKey replays prior results for the same key+input. It is categorised as a Execute tool in the AgentBridge MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on call_action? +

Register the AgentBridge MCP Server MCP server in PolicyLayer and add a rule for call_action: 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 AgentBridge MCP Server. Nothing to install.

What risk level is call_action? +

call_action is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.

Can I rate-limit call_action? +

Yes. Add a rate_limit block to the call_action 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.

How do I block call_action completely? +

Set action: deny in the PolicyLayer policy for call_action. 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.

What MCP server provides call_action? +

call_action is provided by the AgentBridge MCP Server MCP server (marmar9615-cloud/agentbridge-protocol). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.