Index of every page on this site: llms.txt
All skills
wordpress avatar

/wp-playground

@7374e6f official
by wordpresswordpress/agent-skills2.4k stars
351
Run it yourself

Use as the WordPress Playground routing wrapper for ambiguous Playground work, local CLI runs with @wp-playground/cli, playground.wordpress.net share links, browser previews, WebMCP site tools, snapshots, mounts, version switching, and Xdebug. For Blueprint JSON authoring or review, use the blueprint skill directly.

Use this Skill: https://skilld.dev/gh/wordpress/agent-skills/wp-playground

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈83 tokens for metadata: the name and description. ≈861 when used: this file. ≈5.3k more on demand in 6 files.

Description uses 4.2% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈333 characters, or 4.2%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

WordPress Playground

This is a thin routing wrapper. Use it to pick the right Playground workflow, then load only the focused reference or skill needed for the task.

Procedure

  1. Identify the user intent: Blueprint authoring/review, local CLI execution, browser-only website/share link workflow, Xdebug/stuck CLI run, or a mixed Playground request.
  2. Route to the focused source below, loading more than one only when the request has multiple distinct parts.
  3. For mixed requests, delegate Blueprint JSON work to blueprint, then return here for runtime, CLI, debugging, or sharing guidance.
  • Blueprint JSON, schema, steps, resources, bundles, or Blueprint review: use the blueprint skill directly. Do not duplicate Blueprint schema details here.
  • Local CLI execution: read references/cli.md for @wp-playground/cli server, run-blueprint, build-snapshot, mounts, version switching, and local validation.
  • Xdebug or stuck CLI runs: read references/debugging.md for Xdebug, runtime logs, worker flags, and stuck CLI runs.
  • Browser-only Playground website workflows: read references/website.md for Query API and Blueprint URL setup, share links, and browser limitations. It routes existing-site operations to separate WebMCP, Playground MCP, and Sites API references; load only the selected method. Unless the user requests a specific connection method, prefer available WebMCP tools for supported browser operations.

Inputs required

  • The intended workflow: Blueprint authoring, local CLI run, website/share link, snapshot, or debugging.
  • Project or bundle path if local code must be mounted or packaged.
  • Desired WordPress/PHP versions if compatibility matters.
  • Port preference if a local server is needed.
  • Whether browser-only sharing or local filesystem access is required.

Guardrails

  • Playground instances are disposable, SQLite-backed environments; never point them at production data.
  • Keep Blueprint JSON guidance in blueprint so the schema and examples have one source of truth.
  • For local CLI work, verify Node.js 20.18+ and npm/npx before running commands.
  • Browser-only Playground cannot read local filesystem paths; use public URLs, hosted ZIP bundles, or inline Blueprint JSON.

Verification

  • For Blueprint content, validate against the published schema and follow the blueprint skill verification.
  • For local CLI runs, verify the mounted plugin/theme or Blueprint side effects in the Playground instance.
  • For share links, open the generated URL and confirm the expected landing page and installed assets load.

Failure modes

  • Blueprint work routed here: stop and use the blueprint skill for schema keys, steps, resources, bundles, validation, or Blueprint review.
  • Local filesystem needed in a browser-only workflow: use references/cli.md; playground.wordpress.net cannot read local filesystem paths.
  • Shareable browser link requested from a local CLI workflow: use references/website.md; local server URLs are not portable share links.
  • Debugging treated as a second-hop reference: read references/debugging.md directly for Xdebug, logs, worker flags, and stuck CLI runs.

Escalation

  • If the task needs PHP extensions, native database access, persistence, or production-like infrastructure that Playground cannot provide, use a full WordPress stack such as wp-env, Docker, or the project-provided environment.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

2 warnings1mo5 checks · Risk SAFE
  • Gen Agent Trust Hub1mo

    The skill provides a wrapper for WordPress Playground, a developer tool for running WordPress in the browser or via a local CLI. It includes capabilities for executing PHP code, modifying files, and downloading external resources such as Blueprints and plugins. While these are standard features for the intended use case, the ingestion of remote content from arbitrary URLs presents a potential indirect prompt injection surface.

  • Socket1mo

    No alerts

  • Snyk1mo

    Risk: MEDIUM · 1 issue

  • Runlayer9mo

    4/4 files flagged

  • ZeroLeaks7mo

    1 finding · Score: 86/100

Signed by skilld at 7374e6f. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub last week.

Activeupdated 2 months ago
Other metadata
compatibility
Targets WordPress 7.0+, PHP 7.4.0+. Playground CLI requires Node.js 20.18+; runs WordPress in WebAssembly with SQLite.
  • CLI
  • wordpress
  • playground
  • webassembly
  • blueprints
  • php
  • plugin-testing
  • xdebug
  • sqlite

README badge

README badge for wordpress/agent-skills/wp-playground

Spins up disposable WordPress instances locally via Node.js CLI or in the browser, auto-mounting plugins and themes, with support for version switching, blueprints, and Xdebug debugging. Runs WordPress in WebAssembly with SQLite, targeting WordPress 6.9+ and PHP 7.2.24+.

Generated from the current SKILL.md.

Does this work with WordPress versions before 6.9?
No. This skill targets WordPress 6.9 and later, with PHP 7.2.24+.
What Node.js version is required?
Node.js 20.18 or higher is required to run the Playground CLI.
Can I use WordPress Playground with production data?
No. Playground instances are ephemeral and SQLite-backed; you should never point them at production data.
Does this skill support Xdebug debugging?
Yes. You can enable Xdebug with the `--xdebug` flag and connect your IDE (VS Code, PhpStorm) to debug plugin and theme code.
Can I run WordPress Playground entirely in the browser without CLI?
Yes. You can use playground.wordpress.net with blueprint URL fragments or query parameters, or the live Blueprint Editor to author and test without installing the CLI.

Generated from the current SKILL.md. These answers refresh after source changes.