M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
Q

QA Skills MCP Server

Community
Curated QA testing skills directory for AI coding agents
How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #290 of 438
Stars
★ 236 Very popular
Transport
stdio (local process)
Runtime
Node.js
Credentials
No credential needed
License
MIT
Last commit
Tools
6
47FMRS · D

The QA Skills MCP server provides a convenient way to integrate QA testing skills into AI coding agents. It leverages the qaskills.sh directory, offering tools to search and install skills. Official status is not explicitly stated, but it supports multiple agents.

Strongest · Maintenance 12/20 Weakest · Reliability 7/20

Reliability
7/20
Security and permissions
10/20
Maintenance
12/20
Documentation
10/20
Setup experience
8/20
Why each score
Reliability 7/20
Limited evidence: server.json in the package provides a complete init and tool list, but no executable tests or CI evidence confirm the tools actually work as declared. Error handling and edge cases are undocumented. Thus capped at 12, reduced to 7 due to uncertainty.
Security and permissions 10/20
Server uses an environment variable to override API base URL; no hardcoded tokens or secrets found. It makes network requests to external API, but no high-privilege or irreversible operations detected. Permissions and confirmation flows not explicitly documented, but primary risks mitigated. No red lines triggered.
Maintenance 12/20
Repository is active with high stars and MIT license. However, release history and cadence are unclear, dependency updates and security response channel not detailed. Active but governance gaps, hence 12.
Documentation 10/20
README provides clear install and tool list, but lacks parameter specifics, limits, costs, and troubleshooting guides. Examples exist but hidden assumptions. Docs are usable but with gaps.
Setup experience 8/20
Install path is clear: via npx command and client config examples. However, static review lacks executable evidence, reducing score to 8 due to uncertainty.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessWrites / deletes local filesUses the network

Best for

  • QA engineers and SDETs looking to enhance their AI coding agents' testing capabilities.
  • Teams using Claude Code, Cursor, Windsurf, or other MCP-compatible coding agents.
  • Those who want to automate the discovery and installation of testing skills.

Not for

  • Users requiring a transport other than stdio (e.g., remote servers).
  • Those who want to use a private API or self-hosted skills directory unless QASKILLS_API_URL is set.
  • Users who need offline use without internet connection.

Required permissions

  • Access to the qaskills.sh API to search and fetch skill content.
  • Write access to the AI agent's configuration directory to install skills.
  • Optional: Custom API endpoint via QASKILLS_API_URL environment variable.

Risks and side effects

  • Installing skills might modify your AI agent's configuration, potentially causing unexpected behavior.
  • Dependency on the qaskills.sh API may introduce downtime or changes.
  • Skill content might include outdated or inaccurate testing patterns.

Setup

Before you start

Runtime：Node.js

Other optional settings (1)
QASKILLS_API_URL optional Optional env var to override the QASkills API base URL (defaults to https://qaskills.sh); nothing to obtain externally.

For Claude Code: run 'claude mcp add qaskills -- npx -y @qaskills/mcp'. For other MCP clients, add the JSON configuration above to your MCP config.

.mcp.json 
{
  "mcpServers": {
    "qaskills": { "command": "npx", "args": ["-y", "@qaskills/mcp"] }
  }
}

Shown for Claude Code. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "qaskills": {
      "command": "npx",
      "args": [
        "-y",
        "@qaskills/mcp"
      ]
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add qaskills -- npx -y @qaskills/mcp

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

After running claude mcp add qaskills -- npx -y @qaskills/mcp, the client's tool list should show the six tools including search_skills, get_skill, and install_skill; calling list_categories and getting a category list confirms the connection works.

Troubleshooting

  1. Ensure Node.js and npm are installed and up-to-date.
  2. Check network connectivity and API accessibility.
  3. Verify that QASKILLS_API_URL is set correctly if overriding.
  4. Inspect MCP client logs for specific errors.

Things to try

Once connected, you can ask your AI assistant things like:

  • Search for QA skills related to API testing
  • Show me the details of the playwright-e2e skill
  • Install the cypress-e2e skill
  • List the skill categories and show the leaderboard

Tools 6

search_skills read-only
Search for QA testing skills.
get_skill read-only
Get metadata for a skill.
get_skill_content read-only
Get skill content, such as SKILL.md.
install_skill writes
Install a skill into the AI agent's configuration directory.
list_categories read-only
List available skill categories.
get_leaderboard read-only
Get the skill leaderboard.

Use cases

Search for QA testing skills within your IDE or coding agent.
Install testing skills directly into your AI agent without leaving your IDE.
Browse available skill categories and leaderboards.

Supported clients

Claude Code
Cursor
Windsurf

Listed from the project's documentation, not tested by this site.

Overview

The QA Skills MCP server allows you to search, inspect, and install 400+ QA testing skills from qaskills.sh into any MCP client. It provides six tools: search_skills, get_skill, get_skill_content, install_skill, list_categories, and get_leaderboard. This server uses stdio transport and can be configured via the environment variable QASKILLS_API_URL to override the API base URL.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 62.9k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision c9ac503db2ec Data synced 2026-10-12 Read the FMRS scoring method