Agent skill

Ma Sandbox Deploy

by michelangelo-ai in michelangelo-ai/michelangelo

Build a Go service binary, package it into a Docker image, import into k3d, and deploy via helm sync.

Apache-2.0Auto-check passedDevOps & Cloud

Install Ma Sandbox Deploy

skills CLI
$ npx skills add michelangelo-ai/michelangelo --skill ma-sandbox-deploy -a claude-code

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

GitHub CLI
$ gh skill install michelangelo-ai/michelangelo ma-sandbox-deploy --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/michelangelo-ai/michelangelo.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/ma-sandbox-deploy .claude/skills/ma-sandbox-deploy && 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
ma-sandbox-deploy
GitHub stars
118
Token cost
~1.4k tokens
SKILL.md length
245 words
Files
1
Skills in repo
9
Repo updated
First seen
Licence
Apache-2.0

At a glance

Build a Go service binary, package it into a Docker image, import into k3d, and deploy via helm sync.

  • Iterating on apiserver (or another Go service) and wanting to test changes in a running sandbox without waiting for CI
  • SKILL.md covers Full build + deploy sequence, Hot-swap (subsequent iterations), Adapting for controllermgr and Verify what's deployed
  • Calls git, docker and kubectl
  • Tasks that involve Container orchestration

What it does

Ma Sandbox Deploy is an agent skill from michelangelo-ai/michelangelo. Build a Go service binary, package it into a Docker image, import into k3d, and deploy via helm sync. Use when iterating on apiserver (or another Go service) and wanting to test changes in a running sandbox without waiting for CI.

Its SKILL.md is about 1.4k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in DevOps & Cloud, covering Container orchestration and Deployment. It works with Docker. The repository describes itself as: Michelangelo AI: Uber's end-to-end machine learning platform. The licence is Apache-2.0.

When your agent uses it

  • Iterating on apiserver (or another Go service) and wanting to test changes in a running sandbox without waiting for CI
  • Tasks that involve Container orchestration
  • Tasks that involve Deployment

Example prompts

  • “/ma-sandbox-deploy”

Requirements

  • Docker

What it can do on your machine

Read from SKILL.md and the folder at commit 491a9b2. 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

    Shell commands in SKILL.md call:

    • git
    • docker
    • kubectl
    • go

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

  • Network

    No URLs in SKILL.md. Its commands use git, docker and kubectl, which can reach the network depending on how they are called.

    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

Ma Sandbox Deploy loads about 1.4k tokens when it runs. Until then it costs about 62 tokens; SKILL.md has 245 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~62
When it runs · the whole SKILL.md, loaded when a task matches
~1.4k

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 michelangelo-ai/michelangelo at commit 491a9b2, republished under its Apache-2.0 licence (© michelangelo-ai). 245 words, ~1,404 tokens.

Download SKILL.mdSave it as .claude/skills/ma-sandbox-deploy/SKILL.md (or your agent's skills folder).
name
ma-sandbox-deploy
description
Build a Go service binary, package it into a Docker image, import into k3d, and deploy via helm sync. Use when iterating on apiserver (or another Go service) and wanting to test changes in a running sandbox without waiting for CI.
user-invocable
true

Sandbox Local Image Dev Loop

For iterating on Go services (primarily apiserver and controllermgr) inside a running sandbox. Requires a sandbox already running — if not, run ma sandbox create first (see /ma-sandbox-setup).

bash
REPO_ROOT=$(git rev-parse --show-toplevel)

Run this once at the start of your shell session, or prefix paths below with the result.

Full build + deploy sequence

The steps below use apiserver as the example. See Adapting for controllermgr for the controllermgr-specific substitutions.

bash
# 1. Build statically-linked linux/arm64 binary (CGO_ENABLED=0 required — Distroless base has no libc)
cd "$REPO_ROOT/go"
CGO_ENABLED=0 GOOS=linux GOARCH=arm64 go build -o /tmp/apiserver-local ./cmd/apiserver/

# 2. Copy config (only needed once, or when config files change)
cp -r "$REPO_ROOT/go/cmd/apiserver/config" /tmp/apiserver-config

# 3. Prepare a clean build context and build Docker image
mkdir -p /tmp/apiserver-build
cp /tmp/apiserver-local /tmp/apiserver-build/apiserver-local
cp -r /tmp/apiserver-config /tmp/apiserver-build/apiserver-config
docker build \
  -f "$REPO_ROOT/docker/service.Dockerfile" \
  --build-arg BINARY_PATH=apiserver-local \
  --build-arg CONFIG_PATH=apiserver-config \
  --platform linux/arm64 \
  --label git.branch="$(git -C "$REPO_ROOT" rev-parse --abbrev-ref HEAD)" \
  --label git.sha="$(git -C "$REPO_ROOT" rev-parse --short HEAD)" \
  --label git.dirty="$(git -C "$REPO_ROOT" diff --quiet || echo true)" \
  -t michelangelo-apiserver:local \
  /tmp/apiserver-build/

# 4. Import image into k3d cluster
k3d image import michelangelo-apiserver:local -c michelangelo-sandbox

# 5. Sync sandbox with local image override
# IMPORTANT: pass --set images.apiserver=... on EVERY sync call — helm's --reuse-values
# silently reverts to the default ghcr.io image if you omit it
source "$REPO_ROOT/python/.venv/bin/activate"
ma sandbox sync \
  --set images.apiserver=michelangelo-apiserver:local \
  --set images.pullPolicy=IfNotPresent \
  --set ui.enabled=true \
  --set envoy.enabled=true \
  --set controllermgr.enabled=true \
  --set worker.enabled=true

Hot-swap (subsequent iterations)

Check whether config files changed since the last sync:

bash
git diff --name-only HEAD | grep 'cmd/apiserver/config/'

IF that command returns no output — config is unchanged. Skip steps 2–3 (config copy) and run only the binary rebuild + import:

bash
# Rebuild binary, re-import, rollout restart
CGO_ENABLED=0 GOOS=linux GOARCH=arm64 go build -o /tmp/apiserver-local "$REPO_ROOT/go/cmd/apiserver/"
cp /tmp/apiserver-local /tmp/apiserver-build/apiserver-local
docker build \
  -f "$REPO_ROOT/docker/service.Dockerfile" \
  --build-arg BINARY_PATH=apiserver-local \
  --build-arg CONFIG_PATH=apiserver-config \
  --platform linux/arm64 \
  --label git.branch="$(git -C "$REPO_ROOT" rev-parse --abbrev-ref HEAD)" \
  --label git.sha="$(git -C "$REPO_ROOT" rev-parse --short HEAD)" \
  --label git.dirty="$(git -C "$REPO_ROOT" diff --quiet || echo true)" \
  -t michelangelo-apiserver:local \
  /tmp/apiserver-build/
k3d image import michelangelo-apiserver:local -c michelangelo-sandbox
kubectl rollout restart deployment/michelangelo-apiserver
kubectl rollout status deployment/michelangelo-apiserver --timeout=60s

Adapting for controllermgr

Substitute these values in every step above:

apiservercontrollermgr
./cmd/apiserver/./cmd/controllermgr/
/tmp/apiserver-local/tmp/controllermgr-local
/tmp/apiserver-config/tmp/controllermgr-config
/tmp/apiserver-build//tmp/controllermgr-build/
michelangelo-apiserver:localmichelangelo-controllermgr:local
--set images.apiserver=...--set images.controllermgr=...
deployment/michelangelo-apiserverdeployment/michelangelo-controllermgr
app.kubernetes.io/component=apiserverapp.kubernetes.io/component=controllermgr

Check whether cmd/controllermgr/ has a config/ directory — if not, omit steps 2 and the CONFIG_PATH build arg.

Verify what's deployed

The :local tag carries no inherent identity — verify across three signals.

1. It's your local build, not the released image:

bash
kubectl get pod -l app.kubernetes.io/component=apiserver \
  -o jsonpath='{.items[0].spec.containers[0].image}'

Should return michelangelo-apiserver:local, not ghcr.io/michelangelo-ai/apiserver:main.

2. Which branch/commit it was built from (requires the --label flags above):

bash
docker image inspect michelangelo-apiserver:local --format '{{json .Config.Labels}}'
# {"git.branch":"craig.marker/fix-...","git.sha":"d1046e10","git.dirty":"true"}

The labels are baked into the image, so they survive k3d image import — inspecting the local image tells you what the imported (and running) image is.

3. It's your latest build, not a stale pod — compare image build time to pod start time. If the pod started before your last build, you forgot to rollout restart:

bash
docker image inspect michelangelo-apiserver:local --format 'built: {{.Created}}'
kubectl get pod -l app.kubernetes.io/component=apiserver \
  -o jsonpath='started: {.items[0].status.startTime}{"\n"}'

© michelangelo-ai, 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

Just SKILL.md in .claude/skills/ma-sandbox-deploy of michelangelo-ai/michelangelo.

Open the folder on GitHubat commit 491a9b2

Compare with similar skills

Ma Sandbox Deploy 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.

Ma Sandbox Deploy compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Ma Sandbox Deploy this skillmichelangelo-ai/michelangelo118—~1.4kAutomated safety check: PassApache-2.0
LangBot Deployment Guidelangbot-app/LangBot18k—~1.5kAutomated safety check: NotesApache-2.0
Debug Openshell ClusterNVIDIA/OpenShell16k—~20kAutomated safety check: NotesApache-2.0
Aspire MonitoringCommunityToolkit/Aspire627—~3.5kAutomated safety check: PassMIT
Demo Local Rolloutcarverauto/serviceradar921—~4.2kAutomated safety check: PassApache-2.0
Devops EngineerYikai-Liao/symusic1891 repos~1.5kAutomated safety check: PassMIT

Similar skills

  • LangBot Deployment Guide

    langbot-app/LangBot

    Deploys and configures a LangBot instance with Docker Compose or Kubernetes, covering config.yaml, the Box sandbox runtime, the plugin runtime and the global API key.

    18k GitHub stars~1.5k tokensUpdated today
    DevOps & CloudAuto-check: notes
  • Debug Openshell Cluster

    NVIDIA/OpenShell

    Official

    Debug why an OpenShell gateway deployment is unhealthy, unreachable, or unable to create sandboxes.

    16k GitHub stars~20k tokensUpdated yesterday
    DevOps & CloudAuto-check: notes
  • Aspire Monitoring

    CommunityToolkit/Aspire

    ANALYSIS SKILL - Observe Aspire apps: logs, traces, metrics, resource state, telemetry export, browser telemetry, and the standalone dashboard.

    627 GitHub stars~3.5k tokensUpdated 2 days ago
    DevOps & CloudAuto-check passed
  • Demo Local Rollout

    carverauto/serviceradar

    Build unpublished sha-... An agent skill from carverauto/serviceradar.

    921 GitHub stars~4.2k tokensUpdated yesterday
    DevOps & CloudAuto-check passed
  • Devops Engineer

    Yikai-Liao/symusic

    Creates Dockerfiles, configures CI/CD pipelines, writes Kubernetes manifests, and generates Terraform/Pulumi infrastructure templates.

    189 GitHub starsUsed in 1 repo~1.5k tokens
    DevOps & CloudAuto-check passed
  • Deployment

    matrixorigin/memoria

    Deploy Memoria with Docker Compose or Kubernetes. An agent skill from matrixorigin/memoria.

    610 GitHub stars~1.6k tokensUpdated yesterday
    DevOps & CloudAuto-check: notes

More from michelangelo-ai/michelangelo

All 9 skills in this repo
  • Ma Sandbox Debug

    michelangelo-ai/michelangelo

    Tail logs, inspect pods, and diagnose unhealthy services in a running Michelangelo sandbox.

    118 GitHub stars~1k tokensUpdated yesterday
    Auto-check passed
  • Ma Sandbox Setup

    michelangelo-ai/michelangelo

    Canonical setup sequence for the Michelangelo local sandbox.

    118 GitHub stars~936 tokensUpdated yesterday
    Auto-check passed
  • Ma Sandbox Test Plan

    michelangelo-ai/michelangelo

    Build, test, and verify a sandbox change across Go, JS, and Python.

    118 GitHub stars~2k tokensUpdated yesterday
    Auto-check passed
  • Update Docs

    michelangelo-ai/michelangelo

    Update Michelangelo documentation. An agent skill from michelangelo-ai/michelangelo.

    118 GitHub stars~1.6k tokensUpdated yesterday
    Auto-check passed
  • Ma Design Interview

    michelangelo-ai/michelangelo

    Structured interview for designing and implementing changes to the Michelangelo platform.

    118 GitHub stars~2.2k tokensUpdated yesterday
    Auto-check passed
  • Ma Sandbox Reset

    michelangelo-ai/michelangelo

    Tear down the Michelangelo sandbox cluster and recreate it from scratch.

    118 GitHub stars~312 tokensUpdated yesterday
    Auto-check passed

Works with

Categories

Questions about Ma Sandbox Deploy

What does Ma Sandbox Deploy do?

Build a Go service binary, package it into a Docker image, import into k3d, and deploy via helm sync. Ma Sandbox Deploy is an agent skill from michelangelo-ai/michelangelo. Build a Go service binary, package it into a Docker image, import into k3d, and deploy via helm sync.

When should I use Ma Sandbox Deploy?

Ma Sandbox Deploy fits situations like: iterating on apiserver (or another Go service) and wanting to test changes in a running sandbox without waiting for CI; tasks that involve Container orchestration; tasks that involve Deployment.

How do I install Ma Sandbox Deploy in Claude Code?

Run `npx skills add michelangelo-ai/michelangelo --skill ma-sandbox-deploy -a claude-code`. Or copy the skill folder (.claude/skills/ma-sandbox-deploy in michelangelo-ai/michelangelo) into .claude/skills/ma-sandbox-deploy in your project. Claude Code loads it when a task matches its description.

How do I install Ma Sandbox Deploy in Codex?

Run `npx skills add michelangelo-ai/michelangelo --skill ma-sandbox-deploy -a codex`. Or copy the skill folder (.claude/skills/ma-sandbox-deploy in michelangelo-ai/michelangelo) into .agents/skills/ma-sandbox-deploy in your project. Codex loads it when a task matches its description.

Can I use Ma Sandbox Deploy 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 michelangelo-ai/michelangelo --skill ma-sandbox-deploy -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/ma-sandbox-deploy, .gemini/skills/ma-sandbox-deploy, .github/skills/ma-sandbox-deploy and .opencode/skills/ma-sandbox-deploy in your project.

What does Ma Sandbox Deploy need to run?

Going by SKILL.md and its folder, Ma Sandbox Deploy needs the command-line tools its instructions call (git, docker, kubectl and go). Our summary lists: Docker.

Does Ma Sandbox Deploy access the network?

SKILL.md contains no URLs. Its commands use git and docker, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Ma Sandbox Deploy 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 Ma Sandbox Deploy use?

Ma Sandbox Deploy 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 Ma Sandbox Deploy use?

About 1.4k tokens (SKILL.md is roughly 5.6k 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 Ma Sandbox Deploy?

Skills that share tags, products or a category with Ma Sandbox Deploy: LangBot Deployment Guide (langbot-app/LangBot, 18k stars), Debug Openshell Cluster (NVIDIA/OpenShell, 16k stars), Aspire Monitoring (CommunityToolkit/Aspire, 627 stars) and Demo Local Rollout (carverauto/serviceradar, 921 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Ma Sandbox Deploy?

michelangelo-ai (a GitHub organization) maintains it in michelangelo-ai/michelangelo, which has 118 GitHub stars. The repository holds 9 skills in this directory. The repository was last updated on October 10, 2026.

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