Skip to content
Back to skills

Playwright Qa

ASecurity

Playwright QA on Codex with preconfigured test agents and MCP; otherwise use the agent-browser driver.

  • 7 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 10, 2026
ai-agents

Works with

  • mcp

Security analysis

A100/100

Scanned November 10, 2026

npx -y skills add HigorAlves/orc --skill playwright-qa --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Playwright Qa?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Playwright Qa
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/higoralves-playwright-qa-orc/badge)](https://www.skillsdirectory.com/skills/higoralves-playwright-qa-orc)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: playwright-qa
description: Playwright QA on Codex with preconfigured test agents and MCP; otherwise use the agent-browser driver.
---

# Playwright QA on Codex

Before provisioning or installing anything, discover the connected tools and
available Playwright planner, generator, and healer roles. Driver P requires a
Codex-configured `playwright-test` MCP server and all three roles. Claude agent
files and `.mcp.json` do not satisfy this requirement.

If any prerequisite is absent, return `fallback` to the browser-QA driver gate
and offer the bundled agent-browser driver. Do not install dependencies, create
Claude config, run `init-agents`, or invoke `mcp-merge` on this path. The bundled
Codex helper rejects those two provisioning commands before any mutation.

If all prerequisites exist, use `orc-playwright detect <repo>` to locate the
project. If no project exists, return `fallback` before creating files. For a
found project, follow `../browser-qa/SKILL.md` for target/environment gates and
`references/dispatch-prompts.md` for the planner, generator, and healer briefs.
Dispatch through Codex's agent tools with those supplied role instructions.
Read `EXECUTION.md` and follow every shared stage from planning through evidence.
It preserves target credential injection, the guarded-target exclusion of
`@mutating` tests, generated-code scanning, bounded healing, and the packet schema.
Do not run a bare suite command in place of those steps. Report actual failures;
never claim a connected MCP server or role without discovering it.

Automatic Codex agent/MCP provisioning is a documented limitation of this
release. The agent-browser driver remains available for first-time browser QA.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in ai-agents →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.