AI in TestDino for Playwright Testing

TestDino classifies Playwright failures by category, groups related errors across a run, and exposes test data to AI agents over MCP.

TestDino uses AI on your test run data to classify failures and recommend fixes. Connect your own assistants through MCP when you want the same data in chat.

Quick Reference

Feature	Where	Page
Failure classification	Test runs, test cases, dashboard	AI Failure Analysis
Per-test analysis	Test case AI Insights	Test Case AI Insights
Error grouping	Test runs, analytics	Error Grouping
Test Audit	Project AI	AI Test Audit
MCP	Claude, Cursor, ChatGPT	MCP Overview
Agent setup	Docs + prompts	AI Onboarding
Test generation	Recorder, step editor, AI authoring	Test Generation
Playwright skill	Coding agents	Playwright Skill

AI features are on by default. Toggle them in Project Settings → AI Features. Changes apply from the next test run. Providers do not train on customer content: AI data handling.

Failure classification

Every failed test gets a category and confidence score.

Category	Meaning
Actual Bug	Consistent failure that points to a product defect
UI Change	Selector or DOM change broke a test step
Unstable Test	Intermittent failure that passes on retry
Miscellaneous	Setup, data, or CI issue outside the categories above

Classification shows on Test run AI Insights and Test case AI Insights. Correct a misclassification from the feedback form on a test case.

Per-test analysis

Each failing test case gets a category, a confidence score, and fix recommendations drawn from its execution history.

Treat AI recommendations as guidance. Validate before you change product or test code. Learn more in Test Case AI Insights.

Error grouping

AI groups similar errors by message, stack trace, and failure location (assertion, timeout, element not found, network, JavaScript, browser).

Selecting a KPI tile filters the error analysis table. Learn more about clustering in Error Grouping.

MCP and docs for agents

Path	Use when
MCP Overview	The assistant needs live runs, failures, and flaky tests
OpenClaw	You want the same MCP tools from Slack, Telegram, Discord, or WhatsApp
llms.txt / llms-full.txt	The assistant needs docs context only
AI Onboarding	You want the full agent setup path and starter prompt

AI Onboarding

Connect an agent with MCP, CLI, or docs

AI Failure Analysis

Cross-run classification and patterns

TestDino MCP

Query live test data from an assistant

AI Test Audit

Audit suite quality and coverage gaps