Skip to main content
Glama
aliyun-agent-sandbox-mcp

aliyun-agent-sandbox-mcp

Run a shell command

run_command
Destructive

Execute shell commands inside an Alibaba Cloud sandbox and return stdout, stderr, and the exit code; omit sandbox_id to use a temporary sandbox that is removed after the call.

Instructions

Run a shell command inside the sandbox and return stdout, stderr and the exit code. When sandbox_id is omitted a temporary sandbox is created for the call and killed afterwards.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "command": "<command>"
    },
    "name": "run_command"
  }
}

Input Schema

NameRequiredDescriptionDefault
cwdNoWorking directory for the command.
envsNoExtra environment variables for the command.
commandYesShell command to run, e.g. "ls -la /home/user".
sandbox_idNoReuse this sandbox instead of creating a temporary one.
timeout_msNoSandbox TTL and command timeout in milliseconds.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A3.5/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The annotations already declare destructiveHint=true and openWorldHint=true, so the bar is lower, and the description adds the return contract and the fact that an implicit sandbox is killed after the call. It still omits anything about the risk of arbitrary command execution, timeout/disposal of the sandbox on timeout, or permission scope.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences, zero filler, with the core action and return values front-loaded and the sandbox lifecycle note following. Every sentence earns its place.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema, the description correctly covers the return shape (stdout, stderr, exit code), and the ephemeral-sandbox note covers the trickiest parameter interaction. It is still thin for a destructive, open-world execution tool: no mention of timeout consequences, no guidance on error behavior, and no tie-in to the sandbox_* siblings.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so every parameter is already documented and the baseline of 3 applies. The only meaning added beyond the schema is the side effect that an omitted sandbox_id produces a temporary sandbox that is destroyed afterwards; nothing is said about cwd, envs, or timeout_ms.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

"Run a shell command inside the sandbox and return stdout, stderr and the exit code" is a specific verb+resource+output statement, so the agent knows exactly what the call produces. It does not differentiate from the sibling run_code, which is the main remaining ambiguity in purpose.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage through the sandbox_id lifecycle note (omit it to get a throwaway sandbox), which is genuinely useful guidance for choosing between ephemeral and persistent execution. However, it never says when to prefer this over run_code or how it relates to sandbox_create/sandbox_kill, so selection guidance is only implied.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.