start_conversation · Copilot Studio Agent Direct Line MCP

WHAT IT DOES

What start_conversation does on Copilot Studio Agent Direct Line

AI agents invoke start_conversation to trigger actions in Copilot Studio Agent Direct Line. 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 start_conversation needs a policy

While the tool itself does not directly read, write, delete data, or move money, it establishes a session with an external agent system and enables subsequent message-sending operations (as evidenced by sibling tool 'send_message'). This represents an Execute category action because it triggers external system operations whose side effects are contingent on how the agent is configured and what messages follow.

From the tool's definition The tool 'start_conversation' initiates a new conversation session with a Copilot Studio Agent through Direct Line 3.0 API.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

start_conversation Rate-limited

start_conversation 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": {
    "start_conversation": {
      "limits": [
        {
          "counter": "start_conversation_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 Copilot Studio Agent Direct Line, so you can run agents against it safely.

FAQ

Questions about start_conversation

What does the start_conversation tool do? +

Start a new conversation with the Copilot Studio Agent. It is categorised as a Execute tool in the Copilot Studio Agent Direct Line MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on start_conversation? +

Register the Copilot Studio Agent Direct Line MCP server in PolicyLayer and add a rule for start_conversation: 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 Copilot Studio Agent Direct Line. Nothing to install.

What risk level is start_conversation? +

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

Can I rate-limit start_conversation? +

Yes. Add a rate_limit block to the start_conversation 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 start_conversation completely? +

Set action: deny in the PolicyLayer policy for start_conversation. 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 start_conversation? +

start_conversation is provided by the Copilot Studio Agent Direct Line MCP server (swigerb/copilot-studio-agent-direct-line-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.