AI agents invoke new_conversation to trigger actions in Proxima. 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.
Creating a new conversation typically initiates a fresh context or session with an external AI service, which constitutes an Execute action. While not immediately destructive or financial, it represents triggering an external operation whose effects depend on subsequent interactions.
From the tool's definition Tool name 'new_conversation' suggests initiating a new dialogue session with one or more AI providers. Given sibling tools like 'ask_chatgpt', 'ask_claude', etc., this likely triggers external API calls or state changes in the AI gateway.
Documented attack patterns abuse exactly the kind of access new_conversation gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Proxima, and nothing reaches the server without passing your rules. This is the rule we recommend for new_conversation:
{
"version": "1",
"default": "deny",
"tools": {
"new_conversation": {
"limits": [
{
"counter": "new_conversation_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}new_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.
Free to start. No card required.
new_conversation. It is categorised as a Execute tool in the Proxima MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Proxima MCP server in PolicyLayer and add a rule for new_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 Proxima. Nothing to install.
new_conversation 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 new_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.
Set action: deny in the PolicyLayer policy for new_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.
new_conversation is provided by the Proxima MCP server (zen4-bit/proxima). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 50 Proxima tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
50 Proxima tools catalogued and risk-classified — across an index of 42,500+ MCP servers.