Run a bash command in the container. If you've lost track of your environment (chat_id, file URLs, available skills), re-read /home/assistant/README.md. Args: command: Bash command to run in container description: Why I'm running this command Returns: Command output (stdout/stderr)
AI agents invoke bash_tool to trigger actions in Open Computer Use. 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 bash commands with no restrictions mentioned on command scope or impact. In a Docker container context with live browser and network access, bash command execution can trigger external operations, modify system state, access sensitive data, and potentially escape containment.
From the tool's definition Tool name 'bash_tool' with description 'Run a bash command in the container' explicitly permits arbitrary command execution in a Docker workspace with access to terminal, file system, and environment.
Documented attack patterns abuse exactly the kind of access bash_tool gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Open Computer Use, and nothing reaches the server without passing your rules. This is the rule we recommend for bash_tool:
{
"version": "1",
"default": "deny",
"tools": {
"bash_tool": {
"limits": [
{
"counter": "bash_tool_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}bash_tool 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.
Run a bash command in the container. If you've lost track of your environment (chat_id, file URLs, available skills), re-read /home/assistant/README.md. Args: command: Bash command to run in container description: Why I'm running this command Returns: Command output (stdout/stderr). It is categorised as a Execute tool in the Open Computer Use MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Open Computer Use MCP server in PolicyLayer and add a rule for bash_tool: 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 Open Computer Use. Nothing to install.
bash_tool 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 bash_tool 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 bash_tool. 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.
bash_tool is provided by the Open Computer Use MCP server (wide-moat/open-computer-use). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 4 Open Computer Use tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
4 Open Computer Use tools catalogued and risk-classified — across an index of 42,500+ MCP servers.