---
title: "qawolf"
description: "Ask QA Wolf for test coverage, run flows, read results, and drive a cloud browser from your editor."
canonical: https://agentpluginsdirectory.com/plugins/qawolf
last-updated: 2026-09-11
---

# qawolf
Ask QA Wolf for test coverage, run flows, read results, and drive a cloud browser from your editor.
- Slug: qawolf
- Publisher: QA Wolf
- Repository: https://github.com/qawolf/agent-plugins
- Manifest: plugins/qawolf/plugin.json
- Version: 0.1.7
- Category (editorial): other
- Skills: 3 (qawolf-flow-outline, qawolf-onboarding, qawolf)
- MCP servers: 1 (qawolf)
- Stars: 0
- Repository created: 2026-09-04
- Repository last pushed: 2026-09-10
- Publisher type: Organization
- Listing: https://agentpluginsdirectory.com/plugins/qawolf
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What qawolf does, in the publisher's words

Use QA Wolf from your coding agent to request test coverage, run flows, inspect results, and drive a cloud browser. See platform setup for native plugins, portable skills, MCP configuration, and guidance-only limitations.

This is a preview distributed through the QA Wolf plugin repository. It is not yet listed in the providers' official directories. The plugin signs in with OAuth.

QA Wolf clients connect to https://app.qawolf.com/api/mcp and sign in with OAuth through https://signin.qawolf.com. Codex offers the sign-in inside the conversation the first time it calls a QA Wolf tool. The ChatGPT app, Claude Code, Claude Desktop, and Gemini ask when the server is added or on first connection. Check the preview status before use.

- Flow Outline (qawolf-flow-outline) handles every new-flow request. Your coding agent explores through runner computer use, finds context independently, and uses the ask-user tool for gaps. It presents Arrange, Act, Assert outlines for approval, sends approved outlines through agent_send, and monitors creation.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/qawolf/agent-plugins/HEAD/plugins/qawolf/README.md

## Skills

- qawolf-flow-outline: Use whenever the user wants a new QA Wolf flow, test, or end-to-end coverage, even when they already named the journey or supplied a draft outline. Explore the app through runner computer use, resolve gaps, present Arrange-Act-Assert plans for approval, send the approved plans to QA Wolf for implem…
- qawolf-onboarding: Use when the user wants to onboard an application to QA Wolf, get started with QA Wolf, choose their first test, or find the best first flow. Select a useful, achievable onboarding journey, then invoke qawolf-flow-outline for runner exploration, gap questions, AAA approval, creation, and monitoring…
- qawolf: Shared QA Wolf connection, safety, and tool guidance for running flows, investigating failures, managing environments and issues, and driving a cloud browser. Route every new-flow request to qawolf-flow-outline and onboarding or first-flow selection to qawolf-onboarding.

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.

## MCP servers

- qawolf: transport: streamable-http; url: https://app.qawolf.com/api/mcp

Read from the plugin's own mcp.json. Environment variable names only, never values.