Skip to main content
Glama
vndee

LLM Sandbox

by vndee
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Hybrid

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
BACKENDNoThe container backend to use for the sandbox. One of 'docker' (default), 'podman', or 'kubernetes'.
KUBECONFIGNoThe path to your kubeconfig file. Required when using the Kubernetes backend.
DOCKER_HOSTNoThe host to use for the Docker/Podman backend, e.g. 'unix:///var/run/docker.sock' or 'unix:///var/run/podman/podman.sock'. Required if not automatically available in your system.
SANDBOX_CPUSNoCPU limit for the sandbox, normalized to Linux-compatible CPU quota settings.
SANDBOX_MEMORYNoMemory limit for the sandbox, normalized to the backend-safe 'mem_limit' runtime config, e.g. '4g'.
SANDBOX_CAP_DROPNoLinux capabilities to drop for the sandbox container, e.g. 'ALL'.
SANDBOX_CPU_COUNTNoNumber of CPUs for the sandbox, normalized to Linux-compatible CPU quota settings, e.g. '1'.
SANDBOX_MEM_LIMITNoMemory limit for the sandbox runtime config, e.g. '512m'.
SANDBOX_READ_ONLYNoRuntime config to mount the sandbox container as read-only, e.g. 'true'.
SANDBOX_PRIVILEGEDNoWhether to run the sandbox container in privileged mode, e.g. 'true'. Avoid unless explicitly needed.
SANDBOX_NETWORK_MODENoRuntime config for the sandbox network mode (Docker, Podman, and Micromamba backends only). e.g. 'none' to disable egress.
SANDBOX_SECURITY_OPTNoSecurity options for the sandbox container, e.g. 'no-new-privileges:true'.

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}
experimental
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
execute_codeB

Execute code in a secure sandbox environment and automatic visualization capture.

Args:
    code: The code to execute
    language: Programming language (python, javascript, java, cpp, go, r, ruby)
    libraries: List of libraries/packages to install
    timeout: Execution timeout in seconds (default: 30)

Returns:
    List of content items including execution results and any generated visualizations
get_supported_languagesB

Get the list of supported languages.

Returns:
    TextContent: The list of supported languages
get_language_detailsC

Get the details of a language.

Args:
    language: The language to get the details of

Returns:
    TextContent: The details of the language

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription
language_detailsResource containing detailed information about supported languages. Returns: str: The details of the languages
Deploy Server
Shareable, User-Owned Context BackendPeople, applications, and agents read, write, and coordinate together in one context lake. Reach it over MCP, a CLI, a Python SDK, or REST.fulcradynamics.comAd

TDQS

B3.2/5.0

Scored across 3 tools

Disambiguation5/5

The three tools have clearly distinct purposes: execute_code runs code, get_supported_languages lists languages, and get_language_details returns metadata for a specific language. There is no overlap that would cause misselection.

Naming Consistency5/5

All tool names follow a consistent snake_case verb_noun pattern (execute_code, get_supported_languages, get_language_details). The convention is predictable and readable.

Tool Count4/5

Three tools is slightly thin but appropriate for a focused sandbox execution server. Each tool earns its place, though two are lightweight metadata helpers around a single action.

Completeness4/5

The surface covers execution and language introspection well. Minor gaps exist, such as file input/output management or execution status/cancellation, but core workflows are supported.

Maintenance

ActivityActive
ResponsivenessResponsive

Resources