Official agent skill

K8s Agent Sandbox MCP

by kubernetes-sigs in kubernetes-sigs/agent-sandbox

An MCP server skill for managing Kubernetes sandboxes. An agent skill from kubernetes-sigs/agent-sandbox.

OfficialApache-2.0Auto-check passedDevOps & Cloud

Install K8s Agent Sandbox MCP

skills CLI
$ npx skills add kubernetes-sigs/agent-sandbox --skill k8s-agent-sandbox-mcp -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install kubernetes-sigs/agent-sandbox k8s-agent-sandbox-mcp --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/kubernetes-sigs/agent-sandbox.git skills-src && mkdir -p .claude/skills && cp -r skills-src/clients/integrations/mcp-server/skill .claude/skills/k8s-agent-sandbox-mcp && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
k8s-agent-sandbox-mcp
GitHub stars
4.2k
Token cost
~1.3k tokens
SKILL.md length
576 words
Files
3
Skills in repo
7
Repo updated
First seen
Licence
Apache-2.0

At a glance

An MCP server skill for managing Kubernetes sandboxes. An agent skill from kubernetes-sigs/agent-sandbox.

  • Works in 3 steps: Initialize: Call create_sandbox and… → Execute: Call execute_command using the… → Cleanup: Call delete_sandbox when the…
  • Tasks that involve Container orchestration
  • SKILL.md covers Architecture & State, Available MCP Tools and Strict Usage Workflow
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

K8s Agent Sandbox MCP is an agent skill from kubernetes-sigs/agent-sandbox, published by the product's own GitHub organization. An MCP server skill for managing Kubernetes sandboxes. Enables creating, executing commands, managing files, and terminating instances via the official kubernetes-sigs/agent-sandbox MCP server.

Its SKILL.md is about 1.3k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `README.md` and `mcp-config.json`).

It sits in DevOps & Cloud, covering Container orchestration and MCP servers. It works with Kubernetes and Model Context Protocol. The repository describes itself as: agent-sandbox enables easy management of isolated, stateful, singleton workloads, ideal for use cases like AI agent runtimes and reinforcement learning (RL). The licence is Apache-2.0.

When your agent uses it

  • Tasks that involve Container orchestration
  • Tasks that involve MCP servers

Example prompts

  • “/k8s-agent-sandbox-mcp”

Requirements

  • Python 3

Workflow steps

3 steps, taken from the first numbered list in SKILL.md.

  1. Initialize: Call create_sandbox and store the returned sandbox_claim_name in your context.
  2. Execute: Call execute_command using the sandbox_claim_name as many times as needed to complete the task.
  3. Cleanup: Call delete_sandbox when the task is complete, even if previous steps failed. Do not leave orphaned sandboxes running in the…

What it can do on your machine

Read from SKILL.md and the folder at commit d9c1a2b. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

K8s Agent Sandbox MCP loads about 1.3k tokens when it runs. Until then it costs about 54 tokens; SKILL.md has 576 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~54
When it runs · the whole SKILL.md, loaded when a task matches
~1.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from kubernetes-sigs/agent-sandbox at commit d9c1a2b, republished under its Apache-2.0 licence (© kubernetes-sigs). 576 words, ~1,338 tokens.

Download SKILL.mdSave it as .claude/skills/k8s-agent-sandbox-mcp/SKILL.md (or your agent's skills folder). This skill also uses 2 other files; get the full folder from GitHub.
name
k8s-agent-sandbox-mcp
description
An MCP server skill for managing Kubernetes sandboxes. Enables creating, executing commands, managing files, and terminating instances via the official kubernetes-sigs/agent-sandbox MCP server.

Kubernetes Agent Sandbox Manager (MCP)

⚠️ SECURITY WARNING: This MCP server does not have built-in authentication. It exposes tools that can create sandboxes, execute arbitrary commands, and read or write files using the permissions of its Kubernetes service account. Before proceeding, ensure that the server is strictly isolated on a private network or secured behind an authenticating proxy.

Use this skill when a task requires running code, executing untrusted scripts, or performing heavy parallel workloads in an isolated Kubernetes environment. This skill connects to the official kubernetes-sigs/agent-sandbox MCP server. It can be configured using mcp-config.json in this skill's directory.

Architecture & State

Unlike basic shell execution, this sandbox is stateful. When you create a sandbox, a persistent Kubernetes Pod is provisioned and identified by a sandbox_claim_name. You can run multiple sequential commands against the same sandbox_claim_name (e.g., install a package, then run a script). You must retain the sandbox_claim_name value in context for the lifetime of the task.

Available MCP Tools

  • create_sandbox

    • Arguments: warmpool (string), namespace (string), sandbox_ready_timeout (int, optional), labels (dict[string, string], optional), shutdown_after_seconds (int, optional, defaults to 300 — the sandbox self-deletes after 5 minutes unless raised), pod_labels (dict[string, string], optional), pod_annotations (dict[string, string], optional).
    • Returns: JSON object with sandbox_claim_name.
    • Purpose: Create a new sandbox.
  • execute_command

    • Arguments: sandbox_claim_name (string), namespace (string), command (string — shell command or python inline script), timeout (int — seconds before the command times out, optional).
    • Returns: JSON object containing stdout, stderr, and exit_code, or JSON with an error field.
    • Purpose: Executes commands inside the provisioned sandbox.
  • delete_sandbox

    • Arguments: sandbox_claim_name (string), namespace (string).
    • Returns: Success or error JSON object.
    • Purpose: Destroys the Kubernetes Pod and frees resources.
  • download_file

    • Arguments: sandbox_claim_name (string), namespace (string), path (string), binary (bool, optional), timeout (int, optional).
    • Returns: JSON object containing content and bytes_read fields.
    • Purpose: Download a file from a sandbox.
  • file_exists

    • Arguments: sandbox_claim_name (string), namespace (string), path (string), timeout (int, optional).
    • Returns: JSON object containing field exists.
    • Purpose: Check whether a file or directory exists in a sandbox.
  • get_sandbox_status

    • Arguments: sandbox_claim_name (string), namespace (string).
    • Returns: JSON object containing fields status, ready, and message.
    • Purpose: Get the readiness status of a sandbox. Use this before executing commands or transferring files to confirm the sandbox is Ready (e.g. after creation, resume, or warm-pool adoption).
  • list_files

    • Arguments: sandbox_claim_name (string), namespace (string), path (string), timeout (int, optional), max_entries (int, optional, default value is 1000, maximum is 10000).
    • Returns: JSON object containing fields entries, total_entries, and truncated.
    • Purpose: List the contents of a directory in a sandbox. At most max_entries entries are returned (1000 by default). When the directory holds more, the response is truncated and 'truncated' is True while 'total_entries' reports the full count.
  • upload_file

    • Arguments: sandbox_claim_name (string), namespace (string), path (string), content (string), binary (bool, optional), timeout (int, optional).
    • Returns: JSON object containing field bytes_written.
    • Purpose: Upload a file to a sandbox.
Show full SKILL.md (110 more words)Show less

Strict Usage Workflow

ALWAYS follow this exact sequence when using the sandbox:

  1. Initialize: Call create_sandbox and store the returned sandbox_claim_name in your context.
  2. Execute: Call execute_command using the sandbox_claim_name as many times as needed to complete the task.
  3. Cleanup: Call delete_sandbox when the task is complete, even if previous steps failed. Do not leave orphaned sandboxes running in the cluster.
Example Workflow Concept

(Do not write Python wrappers for this, use the provided MCP tools directly)

  1. Tool Call: create_sandbox(warmpool="simple-sandbox-warmpool", namespace="default") → returns {"sandbox_claim_name": "sbx-12345"}
  2. Tool Call: execute_command(sandbox_claim_name="sbx-12345", namespace="default", command="pip install requests")
  3. Tool Call: execute_command(sandbox_claim_name="sbx-12345", namespace="default", command="python -c 'import requests; print(requests.get(\"https://example.com\").status_code)'")
  4. Tool Call: delete_sandbox(sandbox_claim_name="sbx-12345", namespace="default")

© kubernetes-sigs, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 2 other files in clients/integrations/mcp-server/skill of kubernetes-sigs/agent-sandbox.

  • SKILL.md
  • README.md
  • mcp-config.json

Open the folder on GitHubat commit d9c1a2b

Compare with similar skills

K8s Agent Sandbox MCP next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

K8s Agent Sandbox MCP compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
K8s Agent Sandbox MCP this skillkubernetes-sigs/agent-sandbox4.2k—~1.3kAutomated safety check: PassApache-2.0
Kubestellar Consolesickn33/agentic-awesome-skills47k1 repos~1.2kAutomated safety check: PassApache-2.0
Google Cloud Solution Guided Gke AI Migrationgoogle/skills21k—~8.3kAutomated safety check: PassApache-2.0
Eks Best Practicesaws-samples/appmod-blueprints115—~5kAutomated safety check: PassMIT-0
Eks Cost Intelligenceaws-samples/appmod-blueprints115—~3.8kAutomated safety check: WarnMIT-0
Devsydevsy-org/devsy113—~1.7kAutomated safety check: PassMPL-2.0

Similar skills

  • Kubestellar Console

    sickn33/agentic-awesome-skills

    Multi-cluster Kubernetes dashboard with AI-powered operations via MCP server and 10+ built-in agent skills

    47k GitHub starsUsed in 1 repo~1.2k tokens
    DevOps & CloudAuto-check passed
  • Guides the migration of existing AI workloads (Cloud Run, Gemini API, Gemini Enterprise Agent Platform) to self-hosted GKE inference using gcloud and kubectl.

    21k GitHub stars~8.3k tokensUpdated yesterday
    DevOps & CloudAuto-check passed
  • Eks Best Practices

    aws-samples/appmod-blueprints

    Official

    Advisory guidance for Amazon EKS architecture and configuration decisions — compute strategy, networking, security, reliability, cost, autoscaling, observability, multi-tenancy, and upgrade planning.

    115 GitHub stars~5k tokensUpdated 3 days ago
    DevOps & CloudAuto-check passed
  • Eks Cost Intelligence

    aws-samples/appmod-blueprints

    Official

    Run a live EKS cluster cost efficiency assessment — analyze spending across 6 dimensions (compute efficiency, Spot/Graviton adoption, networking, storage, observability, idle resources), calculate a…

    115 GitHub stars~3.8k tokensUpdated 3 days ago
    DevOps & CloudAuto-check: warnings
  • Devsy

    devsy-org/devsy

    Operate Devsy workspaces and providers for end users. An agent skill from devsy-org/devsy.

    113 GitHub stars~1.7k tokensUpdated today
    DevOps & CloudAuto-check passed
  • Dotnet Debugging

    novotnyllc/dotnet-artisan

    Debugs Windows and Linux/macOS applications (native, .NET/CLR, mixed-mode) with WinDbg MCP (crash dumps, !analyze, !syncblk, !dlk, !runaway, !dumpheap, !gcroot, BSOD), dotnet-dump, lldb with SOS…

    232 GitHub stars~2.1k tokensUpdated 3 days ago
    DevelopmentAuto-check passed

More from kubernetes-sigs/agent-sandbox

  • Bump Go Version

    kubernetes-sigs/agent-sandbox

    Official

    Bumps the Go version to the latest release across go.mod, tools.mod, and Dockerfiles.

    4.2k GitHub stars~1.1k tokensUpdated yesterday
    Auto-check passed
  • Fix Flakes

    kubernetes-sigs/agent-sandbox

    Official

    Diagnose and fix flaky tests tracked as open kind/flake issues in kubernetes-sigs/agent-sandbox — reproduce the flake, apply a minimal fix, and open a PR linking the issue.

    4.2k GitHub stars~1.1k tokensUpdated yesterday
    Auto-check passed
  • Test Pyramid

    kubernetes-sigs/agent-sandbox

    Official

    Analyze the repo's unit and E2E tests and propose rebalancing toward a test pyramid — which E2E tests (or assertions inside them) can be covered by unit tests, which unit-level gaps genuinely need…

    4.2k GitHub stars~1.6k tokensUpdated yesterday
    Auto-check passed
  • K8s API Conventions

    kubernetes-sigs/agent-sandbox

    Official

    Guides the agent to follow Kubernetes API conventions for OSS standards.

    4.2k GitHub stars~1.4k tokensUpdated yesterday
    Auto-check passed
  • Dev Rules

    kubernetes-sigs/agent-sandbox

    Official

    Enforces project-specific development rules and conventions.

    4.2k GitHub stars~447 tokensUpdated yesterday
    Auto-check passed
  • Triage Issues

    kubernetes-sigs/agent-sandbox

    Official

    Triage open GitHub issues for kubernetes-sigs/agent-sandbox by mapping them to roadmap.md and assigning k8s priority labels + Kanban Priority (P0–P4) on Project

    4.2k GitHub stars~1.5k tokensUpdated yesterday
    Auto-check passed

Questions about K8s Agent Sandbox MCP

What does K8s Agent Sandbox MCP do?

An MCP server skill for managing Kubernetes sandboxes. An agent skill from kubernetes-sigs/agent-sandbox. K8s Agent Sandbox MCP is an agent skill from kubernetes-sigs/agent-sandbox, published by the product's own GitHub organization. An MCP server skill for managing Kubernetes sandboxes.

When should I use K8s Agent Sandbox MCP?

K8s Agent Sandbox MCP fits situations like: tasks that involve Container orchestration; tasks that involve MCP servers.

How do I install K8s Agent Sandbox MCP in Claude Code?

Run `npx skills add kubernetes-sigs/agent-sandbox --skill k8s-agent-sandbox-mcp -a claude-code`. Or copy the skill folder (clients/integrations/mcp-server/skill in kubernetes-sigs/agent-sandbox) into .claude/skills/k8s-agent-sandbox-mcp in your project. Claude Code loads it when a task matches its description.

How do I install K8s Agent Sandbox MCP in Codex?

Run `npx skills add kubernetes-sigs/agent-sandbox --skill k8s-agent-sandbox-mcp -a codex`. Or copy the skill folder (clients/integrations/mcp-server/skill in kubernetes-sigs/agent-sandbox) into .agents/skills/k8s-agent-sandbox-mcp in your project. Codex loads it when a task matches its description.

Can I use K8s Agent Sandbox MCP in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add kubernetes-sigs/agent-sandbox --skill k8s-agent-sandbox-mcp -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/k8s-agent-sandbox-mcp, .gemini/skills/k8s-agent-sandbox-mcp, .github/skills/k8s-agent-sandbox-mcp and .opencode/skills/k8s-agent-sandbox-mcp in your project.

What does K8s Agent Sandbox MCP need to run?

SKILL.md names no scripts, command-line tools or credentials: K8s Agent Sandbox MCP is instructions for the agent only. Our summary lists: Python 3.

Does K8s Agent Sandbox MCP access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is K8s Agent Sandbox MCP safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does K8s Agent Sandbox MCP use?

K8s Agent Sandbox MCP is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does K8s Agent Sandbox MCP use?

About 1.3k tokens (SKILL.md is roughly 5.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to K8s Agent Sandbox MCP?

Skills that share tags, products or a category with K8s Agent Sandbox MCP: Kubestellar Console (sickn33/agentic-awesome-skills, 47k stars), Google Cloud Solution Guided Gke AI Migration (google/skills, 21k stars), Eks Best Practices (aws-samples/appmod-blueprints, 115 stars) and Eks Cost Intelligence (aws-samples/appmod-blueprints, 115 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains K8s Agent Sandbox MCP?

kubernetes-sigs (a GitHub organization, an official publisher) maintains it in kubernetes-sigs/agent-sandbox, which has 4,210 GitHub stars. The repository holds 7 skills in this directory. The repository was last updated on October 9, 2026.

Source: kubernetes-sigs/agent-sandbox on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.