Official agent skill

Benchmark Sandbox

Run vercel-plugin eval scenarios in Vercel Sandboxes instead of local WezTerm panels.

OfficialCustom licenceAuto-check: notesProduct & Project Management

Install Benchmark Sandbox

skills CLI

$ npx skills add vercel/vercel-plugin --skill benchmark-sandbox -a claude-code

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

GitHub CLI

$ gh skill install vercel/vercel-plugin benchmark-sandbox --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/vercel/vercel-plugin.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/benchmark-sandbox .claude/skills/benchmark-sandbox && 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/

Paste into Claude Code

Install the "benchmark-sandbox" agent skill from https://github.com/vercel/vercel-plugin/tree/main/.claude/skills/benchmark-sandbox into .claude/skills/benchmark-sandbox/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "benchmark-sandbox", then confirm the skill loads.

Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

$skill-installer

$skill-installer install https://github.com/vercel/vercel-plugin/tree/main/.claude/skills/benchmark-sandbox

Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.

skills CLI

$ npx skills add vercel/vercel-plugin --skill benchmark-sandbox -a codex

Project install goes to .agents/skills/; add -g for ~/.codex/skills/.

GitHub CLI

$ gh skill install vercel/vercel-plugin benchmark-sandbox --agent codex

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/vercel/vercel-plugin.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.claude/skills/benchmark-sandbox .agents/skills/benchmark-sandbox && rm -rf skills-src

Use ~/.agents/skills/ instead of .agents/skills for a personal install.

Codex skills documentation · loads skills from .agents/skills/

Paste into Codex

Install the "benchmark-sandbox" agent skill from https://github.com/vercel/vercel-plugin/tree/main/.claude/skills/benchmark-sandbox into .agents/skills/benchmark-sandbox/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "benchmark-sandbox", then confirm the skill loads.

Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add vercel/vercel-plugin --skill benchmark-sandbox -a cursor

Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.

GitHub CLI

$ gh skill install vercel/vercel-plugin benchmark-sandbox --agent cursor

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/vercel/vercel-plugin.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.claude/skills/benchmark-sandbox .cursor/skills/benchmark-sandbox && rm -rf skills-src

Use ~/.cursor/skills/ instead of .cursor/skills for a personal install.

Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/

Paste into Cursor

Install the "benchmark-sandbox" agent skill from https://github.com/vercel/vercel-plugin/tree/main/.claude/skills/benchmark-sandbox into .cursor/skills/benchmark-sandbox/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "benchmark-sandbox", then confirm the skill loads.

Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

gemini skills install

$ gemini skills install https://github.com/vercel/vercel-plugin.git --path .claude/skills/benchmark-sandbox

--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.

skills CLI

$ npx skills add vercel/vercel-plugin --skill benchmark-sandbox -a gemini-cli

Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.

GitHub CLI

$ gh skill install vercel/vercel-plugin benchmark-sandbox --agent gemini-cli

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/vercel/vercel-plugin.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.claude/skills/benchmark-sandbox .gemini/skills/benchmark-sandbox && rm -rf skills-src

Use ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.

Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/

Paste into Gemini CLI

Install the "benchmark-sandbox" agent skill from https://github.com/vercel/vercel-plugin/tree/main/.claude/skills/benchmark-sandbox into .gemini/skills/benchmark-sandbox/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "benchmark-sandbox", then confirm the skill loads.

Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

GitHub CLI

$ gh skill install vercel/vercel-plugin benchmark-sandbox

Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).

skills CLI

$ npx skills add vercel/vercel-plugin --skill benchmark-sandbox -a github-copilot

Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.

Manual copy

$ git clone --depth 1 https://github.com/vercel/vercel-plugin.git skills-src && mkdir -p .github/skills && cp -r skills-src/.claude/skills/benchmark-sandbox .github/skills/benchmark-sandbox && rm -rf skills-src

Use ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.

GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/

Paste into GitHub Copilot

Install the "benchmark-sandbox" agent skill from https://github.com/vercel/vercel-plugin/tree/main/.claude/skills/benchmark-sandbox into .github/skills/benchmark-sandbox/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "benchmark-sandbox", then confirm the skill loads.

GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add vercel/vercel-plugin --skill benchmark-sandbox -a opencode

OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.

GitHub CLI

$ gh skill install vercel/vercel-plugin benchmark-sandbox --agent opencode

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/vercel/vercel-plugin.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.claude/skills/benchmark-sandbox .opencode/skills/benchmark-sandbox && rm -rf skills-src

Use ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.

OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/

Paste into OpenCode

Install the "benchmark-sandbox" agent skill from https://github.com/vercel/vercel-plugin/tree/main/.claude/skills/benchmark-sandbox into .opencode/skills/benchmark-sandbox/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "benchmark-sandbox", then confirm the skill loads.

OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

Facts

Skill name: benchmark-sandbox
GitHub stars: 301
Token cost: ~5.4k tokens
SKILL.md length: 1,946 words
Files: 9
Skills in repo: 38
Repo updated: Oct 6, 2026
First seen: Oct 7, 2026
Licence: Custom licence

At a glance

Run vercel-plugin eval scenarios in Vercel Sandboxes instead of local WezTerm panels.

Works in 12 steps: Snapshots work: sandbox.snapshot()… → Plugin install: Use npx add-plugin -s… → File uploads: Use sandbox.writeFiles([{… → …
Tasks that involve Test coverage
SKILL.md covers Proven Working Script, Dynamic Scenarios (Recommended…, Structured Scoring (Haiku) and Critical Sandbox Environment…, plus 12 more sections
Runs TypeScript scripts from its folder; calls bun, npx and vercel; reaches sb-xxx.vercel.run and xxx.vercel.app; needs ANTHROPIC_API_KEY and ANTHROPIC_AUTH_TOKEN

What it does

Benchmark Sandbox is an agent skill from vercel/vercel-plugin, published by the product's own GitHub organization. Run vercel-plugin eval scenarios in Vercel Sandboxes instead of local WezTerm panels. Provisions ephemeral microVMs with Claude Code + plugin pre-installed, runs benchmark prompts, extracts hook artifacts, and produces coverage reports.

Its SKILL.md is about 5.4k tokens, which your agent loads only when the skill is triggered. The skill folder holds 9 other files (for example `logger.ts`, `run-eval.ts` and `sandbox-analyze.ts`).

It sits in Product & Project Management, covering Test coverage. It works with Vercel. The repository describes itself as: Comprehensive Vercel ecosystem plugin — relational knowledge graph, skills for every major product, specialized agents, and Vercel conventions. Turns any AI agent into a Vercel…

When your agent uses it

Tasks that involve Test coverage

Example prompts

“/benchmark-sandbox”

Requirements

Node.js
A credential in ANTHROPIC_AUTH_TOKEN
A credential in ANTHROPIC_API_KEY

Workflow steps

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

Snapshots work: sandbox.snapshot() preserves files AND npm globals. Use it after build to create a restore point before verify/deploy…
Plugin install: Use npx add-plugin -s project -y --target claude-code — works because claude is in PATH after npm install -g. The --target…
File uploads: Use sandbox.writeFiles([{ path, content: Buffer }]) — NOT runCommand heredocs. Heredocs with special characters cause 400…
Claude flags: Always use --dangerously-skip-permissions --debug. The --debug flag writes to ~/.claude/debug/.
Auth: API key from macOS Keychain (ANTHROPIC_AUTH_TOKEN — a vck_* Vercel Claude Key for AI Gateway), Vercel token from…
OIDC for sandbox SDK: Run npx vercel link --scope vercel-labs -y + npx vercel env pull once before first use.
Port exposure: Pass ports: [3000] in Sandbox.create() to get a public URL immediately via sandbox.domain(3000). Works on v1.8.0 — URL is…
extendTimeout: Use sandbox.extendTimeout(ms) to keep sandboxes alive past their initial timeout. Verified working — extends by the…
Background commands: runCommand with backgrounded processes (& or nohup) may throw ZodError on v1. Write a script file first, then execute…
Session cleanup race: The session-end-cleanup.mjs hook deletes /tmp/vercel-plugin-*-seen-skills.d/ on session end. Extract artifacts…
agent-browser works in sandboxes: Install via npm install -g agent-browser. Claude Code can use it for browser-based verification inside…
No hobby tier cap: Early 301s timeouts were from lower default timeout values in earlier script iterations, not a tier limitation…

What it can do on your machine

Read from SKILL.md and the folder at commit 82fa491. 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
Ships script files (TypeScript), which the agent can run.
Shell commands in SKILL.md call:
- bun
- npx
- vercel
- claude
- npm
- sh
From the folder's file list and the shell code blocks in SKILL.md.
Network
Hosts in commands or code, which the agent is likely to contact:
- sb-xxx.vercel.run
- xxx.vercel.app
- ai-gateway.vercel.sh
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names these keys or tokens, usually read from environment variables:
- ANTHROPIC_API_KEY
- ANTHROPIC_AUTH_TOKEN
- VERCEL_TOKEN
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Benchmark Sandbox loads about 5.4k tokens when it runs. Until then it costs about 64 tokens; SKILL.md has 1,946 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it: ~64
When it runs · the whole SKILL.md, loaded when a task matches: ~5.4k

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: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

NoteMentions a .env fileSKILL.md:278
```
npx vercel env pull .env.local
```

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

Its licence (Custom licence) doesn't allow us to republish the file, so here is its outline and opening line. It has 1,946 words (~5,403 tokens).

“Run benchmark scenarios inside Vercel Sandboxes — ephemeral Firecracker microVMs with node24. Each sandbox gets a fresh Claude Code + Vercel CLI + agent-browser install, the local vercel-plugin uploaded, and runs a 3-phase eval pipeline:”

— opening of SKILL.md by vercel, Custom licence

name: benchmark-sandbox

Outline of SKILL.md

Benchmark Sandbox — Remote Eval via Vercel Sandboxes
Proven Working Script
CLI Flags
Dynamic Scenarios (Recommended Approach)
Scenario JSON Format
Prompt Design Guidelines
Structured Scoring (Haiku)
Build Score Schema
Verify Score Schema (per user story)
Deploy Score Schema
Critical Sandbox Environment Facts
Key Discoveries (Hard-Won)
When to Use This vs benchmark-agents
How It Works
Sandbox Session Flow (Per Scenario)
Verification Phase Details
Deploy Phase Details
DO NOT (Hard Rules)
Prerequisites
Commands
Run eval with dynamic scenarios (recommended)
Monitoring While Running
Artifact Export Layout
Proven Results (2026-03-10)
Known Limitations

Read the full SKILL.md on GitHub

Files

SKILL.md and 8 other files in .claude/skills/benchmark-sandbox of vercel/vercel-plugin.

SKILL.md
logger.ts
run-eval.ts
sandbox-analyze.ts
sandbox-runner.ts
spike/create-snapshot.ts
spike/interactive-proof.ts
spike/provision.ts
types.ts

Open the folder on GitHubat commit 82fa491

Compare with similar skills

Benchmark Sandbox 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.

Benchmark Sandbox compared with similar skills
Skill	Stars	Used in	Tokens	Auto-check	Licence	Repo updated
Benchmark Sandbox this skillvercel/vercel-plugin	301	—	~5.4k	Automated safety check: Notes	Custom licence	today
Enrich Release Roadmaptrickle-labs/pg-trickle	146	—	~1.9k	Automated safety check: Pass	Apache-2.0	today
AlignmentOvid/paad	131	—	~4.9k	Automated safety check: Pass	MIT	yesterday
Nextjs Supabase VercelTechNomadCode/AI-Product-Development-Toolkit	1k	1 repos	~159	Automated safety check: Pass	MIT	yesterday
Test Planningpetrkindlmann/qa-skills	163	—	~4.2k	Automated safety check: Pass	MIT	3 mo ago
Sveltekit WebappLeoYeAI/openclaw-master-skills	2.2k	—	~5.2k	Automated safety check: Notes	MIT	2 mo ago

Similar skills

Enrich Release Roadmap
trickle-labs/pg-trickle
Enrich a pgtrickle release roadmap with prioritised items across six quality pillars: correctness, stability, performance, scalability, ease-of-use, and test coverage.
146 GitHub stars~1.9k tokensUpdated today
Product & Project ManagementAuto-check passed
Alignment
Ovid/paad
A skill your agent uses when verifying that requirements/specs/PRDs and their implementation plans match — before starting work, after a spec or plan update, or when suspecting coverage gaps, scope…
131 GitHub stars~4.9k tokensUpdated yesterday
Product & Project ManagementAuto-check passed
Nextjs Supabase Vercel
TechNomadCode/AI-Product-Development-Toolkit
Set up the current empty folder with the AI App Starter, so the agent builds a Next.js, Supabase and Vercel app with the owner step by step, from idea to launch.
1k GitHub starsUsed in 1 repo~159 tokens
Product & Project ManagementAuto-check passed
Test Planning
petrkindlmann/qa-skills
Build a single sprint or release test plan. An agent skill from petrkindlmann/qa-skills.
163 GitHub stars~4.2k tokensUpdated 3 mo ago
Testing & QAAuto-check passed
Sveltekit Webapp
LeoYeAI/openclaw-master-skills
Scaffold and configure a production-ready SvelteKit PWA with opinionated defaults.
2.2k GitHub stars~5.2k tokensUpdated 2 mo ago
Product & Project ManagementAuto-check: notes
Vibe Scenario Matrix
ash1794/vibe-engineering
Generates a behavioral scenario acceptance matrix for comprehensive test coverage planning.
162 GitHub stars~663 tokensUpdated yesterday
Testing & QAAuto-check passed

Questions about Benchmark Sandbox

What does Benchmark Sandbox do?

Run vercel-plugin eval scenarios in Vercel Sandboxes instead of local WezTerm panels. Benchmark Sandbox is an agent skill from vercel/vercel-plugin, published by the product's own GitHub organization. Run vercel-plugin eval scenarios in Vercel Sandboxes instead of local WezTerm panels.

When should I use Benchmark Sandbox?

Benchmark Sandbox fits situations like: tasks that involve Test coverage.

How do I install Benchmark Sandbox in Claude Code?

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

How do I install Benchmark Sandbox in Codex?

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

Can I use Benchmark Sandbox 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 vercel/vercel-plugin --skill benchmark-sandbox -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/benchmark-sandbox, .gemini/skills/benchmark-sandbox, .github/skills/benchmark-sandbox and .opencode/skills/benchmark-sandbox in your project.

What does Benchmark Sandbox need to run?

Going by SKILL.md and its folder, Benchmark Sandbox needs TypeScript for the scripts in its folder, the command-line tools its instructions call (bun, npx, vercel, claude, npm and sh) and credentials named ANTHROPIC_API_KEY, ANTHROPIC_AUTH_TOKEN and VERCEL_TOKEN. Our summary lists: Node.js; A credential in ANTHROPIC_AUTH_TOKEN; A credential in ANTHROPIC_API_KEY.

Does Benchmark Sandbox access the network?

SKILL.md names 3 domains. In commands or code: sb-xxx.vercel.run, xxx.vercel.app and ai-gateway.vercel.sh; the agent is likely to contact these when it follows the instructions. This is read from the text; nothing was executed.

Is Benchmark Sandbox safe to install?

Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Benchmark Sandbox use?

Benchmark Sandbox has a licence file (the repository's licence) that doesn't match a standard licence. Read it on GitHub before reusing the skill.

How many tokens does Benchmark Sandbox use?

About 5.4k tokens (SKILL.md is roughly 22k 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 Benchmark Sandbox?

Skills that share tags, products or a category with Benchmark Sandbox: Enrich Release Roadmap (trickle-labs/pg-trickle, 146 stars), Alignment (Ovid/paad, 131 stars), Nextjs Supabase Vercel (TechNomadCode/AI-Product-Development-Toolkit, 1k stars) and Test Planning (petrkindlmann/qa-skills, 163 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Benchmark Sandbox?

vercel (a GitHub organization, an official publisher) maintains it in vercel/vercel-plugin, which has 301 GitHub stars. The repository holds 38 skills in this directory. The repository was last updated on October 6, 2026.

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

Benchmark Sandbox

Install Benchmark Sandbox

Facts

At a glance

What it does

When your agent uses it

Example prompts

Requirements

Workflow steps

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Context cost

Safety

SKILL.md

Files

Compare with similar skills

Similar skills

Enrich Release Roadmap

Alignment

Nextjs Supabase Vercel

Test Planning

Sveltekit Webapp

Vibe Scenario Matrix

More from vercel/vercel-plugin

Deployments Cicd

Plugin Audit

Vercel CLI

AI Gateway

Vercel Services

Access Protected Vercel Deployment

Questions about Benchmark Sandbox

Install Benchmark Sandbox

Facts

At a glance

What it does

When your agent uses it

Example prompts

Requirements

Workflow steps

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Context cost

Safety

SKILL.md

Files

Compare with similar skills

Similar skills

More from vercel/vercel-plugin

Related

Questions about Benchmark Sandbox