Dispatch browser and UI test work to the specialized testing agent that matches the requested outcome.
3 stars
0 votes
0 copies
0 views
Added September 29, 2026
ai-agentstestingperformance
Works with
mcp
Security analysis
A100/100
Scanned September 29, 2026
$npx -y skills add dryvist/claude-code-plugins --skill test --agent claude-code
Installs into .claude/skills of the current project.
Are you the author of Test?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/dryvist-test)
---
name: test
description: Dispatch browser and UI test work to the specialized testing agent that matches the requested outcome.
---
# test
Choose one testing agent, then run it as a subagent with the target, mode, and
acceptance criteria in its prompt:
| Need | Agent |
| --- | --- |
| Run deterministic Playwright checks or inspect screenshots | `ui-smoke` |
| Design or update a spec | `spec-author` |
| Classify a failed check and repair a stale selector | `healer` |
| Inspect browser performance, console, or network behavior | `perf-debug` |
| Explore a browser flow without deciding a gate | `explorer` |
Use the testing-agents plugin for the selected agent. Keep deterministic
Playwright results as the gate; `explorer` uses Browser Use only for
non-gating investigation.
`perf-debug` runs in its dedicated local-agent environment. Claude plugin
agents ignore `mcpServers` frontmatter, so do not expect this dispatcher to
load Chrome DevTools MCP tools into the main session.
