PluginBench
MCP Server
Maintained
MIT

Exec Sandbox MCP Server

io.github.inhuman/mcp-exec

RepositoryRegistry
How to choose

What this kind of tool is for

  • Automate a build, test, or deployment step you repeat by hand.
  • Review or refactor code in an existing repository without leaving your agent.
  • Bridge an internal service or CLI into the agent toolset.

How to compare servers like this

Scope of control:
Repos and CI are hard to undo; prefer tools that ask before anything irreversible.
Environment fit:
Match the runtime it expects (Node, Python, a container) to what you actually run.
Maintenance signal:
Check the last-commit date — dev tooling that lags upstream breaks first.
Config surface:
Prefer tools whose environment-variable or config keys are documented plainly.

This page carries your copy-paste config: pick your MCP client in the install section on this page and use the copy button.

What is the Exec Sandbox MCP server?

Single exec tool running caller Python in a network-isolated locked-down sandbox.

How to install Exec Sandbox

Copy-paste configuration for popular MCP clients.

transport: stdio
Config generated by PluginBench — verify against the source before use.
Environment / auth
  • MCP_EXEC_TRANSPORT

    MCP transport. Must be 'stdio' for direct docker/stdio use.

  • MCP_EXEC_DEFAULT_TIMEOUT_S

    Default wall-clock timeout per exec, seconds.

  • MCP_EXEC_MAX_TIMEOUT_S

    Maximum wall-clock timeout a caller may request, seconds.

  • MCP_EXEC_MAX_OUTPUT_BYTES

    Cap on combined stdout/stderr before truncation.

  • MCP_EXEC_MAX_STDIN_BYTES

    Cap on stdin passed to the sandboxed code.

  • MCP_EXEC_AUTH_TOKEN
    secret

    Optional X-MCP-AUTH token (http/sse transports only).

~/Library/Application Support/Claude/claude_desktop_config.json
JSON config
{
  "mcpServers": {
    "mcp-exec": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "docker.io/idconstruct/mcp-exec:v0.5.4",
        "--network",
        "none",
        "--read-only",
        "--cap-drop",
        "ALL",
        "--security-opt",
        "no-new-privileges",
        "--user",
        "65532:65532",
        "--tmpfs",
        "/tmp:rw,noexec,nosuid,size=64m",
        "--memory",
        "256m",
        "--pids-limit",
        "128",
        "--cpus",
        "1"
      ],
      "env": {
        "MCP_EXEC_TRANSPORT": "<YOUR_MCP_EXEC_TRANSPORT>",
        "MCP_EXEC_DEFAULT_TIMEOUT_S": "<YOUR_MCP_EXEC_DEFAULT_TIMEOUT_S>",
        "MCP_EXEC_MAX_TIMEOUT_S": "<YOUR_MCP_EXEC_MAX_TIMEOUT_S>",
        "MCP_EXEC_MAX_OUTPUT_BYTES": "<YOUR_MCP_EXEC_MAX_OUTPUT_BYTES>",
        "MCP_EXEC_MAX_STDIN_BYTES": "<YOUR_MCP_EXEC_MAX_STDIN_BYTES>",
        "MCP_EXEC_AUTH_TOKEN": "<YOUR_MCP_EXEC_AUTH_TOKEN>"
      }
    }
  }
}

Related MCP servers

Runs a command in a throwaway Kubernetes pod and returns exit code, output and artifacts.

0
Go
MIT
View repository →
SSSSH Fleet logo

SSH Fleet

Maintained

Curated ssh_probe diagnostics and single-host ssh_exec over an allowlisted SSH fleet.

0
Go
MIT
View repository →

AI-powered Git repository analysis for coding agents — understand any codebase instantly with ground truth.

18
TypeScript
MIT
View repository →

Turn an IDF domain into an MCP server with invariants and role scopes in every tool description.

0
JavaScript
MIT
View repository →
INInite Brain logo

Open-source bitemporal memory layer for LLM agents with knowledge graphs and conflict resolution.

32
TypeScript
AGPL-3.0
View repository →

Deterministic reasoning stack for AI agents: simulate, decide & compute, plus cross-domain tools.

0
JavaScript
MIT
View repository →