run_pipeline · Gst MCP: Risk & Policy

WHAT IT DOES

What run_pipeline does on Gst

AI agents invoke run_pipeline to trigger actions in Gst. 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_pipeline needs a policy

This tool executes external operations (GStreamer pipelines) whose effects depend entirely on the pipeline arguments provided. While GStreamer is primarily a media framework, pipelines can be constructed to perform side effects including file creation/modification, network access, and system operations.

From the tool's definition Tool description states 'Run a GStreamer pipeline' with synchronous and asynchronous execution modes. GStreamer pipelines can perform arbitrary media processing, file I/O, network operations, and invoke external processes depending on pipeline configuration.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

run_pipeline Rate-limited

run_pipeline 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_pipeline": {
      "limits": [
        {
          "counter": "run_pipeline_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 Gst, so you can run agents against it safely.

FAQ

Questions about run_pipeline

What does the run_pipeline tool do? +

Run a GStreamer pipeline. Can run synchronously with timeout or asynchronously. CRITICAL: Do NOT call this tool unless the user has EXPLICITLY confirmed they want to run the pipeline. Always use validate_pipeline first to check and present the pipeline, then wait for the user to say they want to run it before calling this tool. IMPORTANT: Always provide working_directory so output files are created in the user. It is categorised as a Execute tool in the Gst 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_pipeline? +

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

What risk level is run_pipeline? +

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

Can I rate-limit run_pipeline? +

Yes. Add a rate_limit block to the run_pipeline 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_pipeline completely? +

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

run_pipeline is provided by the Gst MCP server (wizenink/gst-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.