run_anomaly_detection · Razorpay Recon MCP: Risk & Policy

WHAT IT DOES

What run_anomaly_detection does on Razorpay Recon

AI agents invoke run_anomaly_detection to trigger actions in Razorpay Recon. 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 run_anomaly_detection needs a policy

While the primary function (anomaly detection on payment data) is analytical and read-focused, the tool's documented behavior includes triggering WhatsApp notifications to external parties (finance team). This outbound communication triggered by the tool's execution moves it beyond Read into Execute category.

From the tool's definition The tool 'run_anomaly_detection' executes a data analysis operation with external side effects: it 'Sends WhatsApp alerts to the finance team for high-severity anomalies.' This is a triggered external action dependent on the tool's execution and the data…

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

run_anomaly_detection Rate-limited

run_anomaly_detection 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": {
    "run_anomaly_detection": {
      "limits": [
        {
          "counter": "run_anomaly_detection_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 Razorpay Recon, so you can run agents against it safely.

FAQ

Questions about run_anomaly_detection

What does the run_anomaly_detection tool do? +

Run anomaly detection on recent Razorpay data. Checks for: failed payments >₹5,000, unmatched settlements, and refund volume spikes. Sends WhatsApp alerts to the finance team for high-severity anomalies. Args: days: Number of days to look back for payment data (default: 1). Returns: List of detected anomalies with type, severity, and details. It is categorised as a Execute tool in the Razorpay Recon MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on run_anomaly_detection? +

Register the Razorpay Recon MCP server in PolicyLayer and add a rule for run_anomaly_detection: 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 Razorpay Recon. Nothing to install.

What risk level is run_anomaly_detection? +

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

Can I rate-limit run_anomaly_detection? +

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

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

run_anomaly_detection is provided by the Razorpay Recon MCP server (qalalabs/razorpay-recon-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.