compute · Scicompute MCP: Risk & Policy

WHAT IT DOES

What compute does on Scicompute

AI agents invoke compute to trigger actions in Scicompute. 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 compute needs a policy

This tool executes code in multiple scientific computing environments. While the outputs are typically mathematical/visualization results rather than system-level operations, code execution across these backends carries significant risk: arbitrary computations could consume resources, access sensitive data, produce incorrect results used in critical decisions, or (depending on backend configurations) potentially…

From the tool's definition Tool description states 'Execute scientific computing code' with support for multiple computational backends (Mathematica, Sage, Python, R, Octave, Julia, Maxima).

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

compute Rate-limited

compute 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": {
    "compute": {
      "limits": [
        {
          "counter": "compute_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 Scicompute, so you can run agents against it safely.

FAQ

Questions about compute

What does the compute tool do? +

Execute scientific computing code. Supports multiple backends (mathematica, sage, py_scientific, r, octave, julia, maxima). Use. It is categorised as a Execute tool in the Scicompute MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on compute? +

Register the Scicompute MCP server in PolicyLayer and add a rule for compute: 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 Scicompute. Nothing to install.

What risk level is compute? +

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

Can I rate-limit compute? +

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

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

compute is provided by the Scicompute MCP server (sanshanjianke/scicompute-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.