run_paragraph · Zeppelin MCP: Risk & Policy

WHAT IT DOES

What run_paragraph does on Zeppelin

AI agents invoke run_paragraph to trigger actions in Zeppelin. 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_paragraph needs a policy

This tool executes code within Zeppelin paragraphs. While it doesn't inherently delete or move money, code execution is a core Execute category action with significant blast radius—an agent could run destructive queries, access sensitive data, modify systems, or trigger unintended side effects depending on what code is in the paragraph.

From the tool's definition Tool name 'run_paragraph' and description 'Execute a single paragraph' directly indicate code execution. The description states it runs a paragraph synchronously, which in Apache Zeppelin means executing potentially arbitrary code (Scala, Python, SQL, etc.)…

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

run_paragraph Rate-limited

run_paragraph 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_paragraph": {
      "limits": [
        {
          "counter": "run_paragraph_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 Zeppelin, so you can run agents against it safely.

FAQ

Questions about run_paragraph

What does the run_paragraph tool do? +

Execute a single paragraph and return its output. This is a synchronous call that waits for execution to complete. It is categorised as a Execute tool in the Zeppelin 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_paragraph? +

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

What risk level is run_paragraph? +

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

Can I rate-limit run_paragraph? +

Yes. Add a rate_limit block to the run_paragraph 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_paragraph completely? +

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

run_paragraph is provided by the Zeppelin MCP server (mihneamanolache/zeppelin-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.