AI agents invoke run to trigger actions in Bash. 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.
This tool allows execution of arbitrary shell commands, which is the definition of Execute category. Severity is critical because shell command execution has maximal blast radius - an AI agent could delete files, exfiltrate data, modify system state, launch attacks, or compromise the entire system depending on the agent's permissions.
From the tool's definition "Execute a shell command and return output" - the tool directly runs arbitrary shell commands with no restrictions mentioned.
Documented attack patterns abuse exactly the kind of access run gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Bash, and nothing reaches the server without passing your rules. This is the rule we recommend for run:
{
"version": "1",
"default": "deny",
"tools": {
"run": {
"limits": [
{
"counter": "run_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}run 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.
Execute a shell command and return output. It is categorised as a Execute tool in the Bash MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Bash MCP server in PolicyLayer and add a rule for run: 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 Bash. Nothing to install.
run 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 run 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 run. 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.
run is provided by the Bash MCP server (tinywind/bash-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Bash, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
4 Bash tools catalogued and risk-classified — across an index of 43,000+ MCP servers.