Agentic Testing: Where Agents Fit in the E2E Testing Stack
Slack ran 200+ agentic E2E tests. Here's what actually works—and what doesn't replace deterministic testing.
Why it matters
As AI agents move into production QA workflows, engineering leaders need practical guidance on where agents add value vs. where traditional testing remains essential. Slack's large-scale experiment provides rare empirical data on agent-driven testing at scale.
The key facts
11 to know200+ agentic E2E workflows tested
Used Playwright MCP, Playwright CLI, and agent-generated tests
Tests run on non-production data in isolated workspaces
Agentic testing positioned as exploratory layer, not replacement for deterministic tests
Published by Slack Engineering (credible first-party data)
Focus on practical integration into existing E2E testing stack
200+ agentic E2E test workflows executed
Tested using Playwright MCP, Playwright CLI, agent-generated Playwright tests
Non-production test environment
Focus: exploratory layer vs. deterministic test replacement
Slack engineering case study on agent integration in QA stack
Go to the source
Slack Engineeringslack.engineering
Publisher excerpt: Abstract Agent-driven end-to-end (E2E) tests add a new exploratory layer to testing, but should they replace traditional deterministic tests? We ran more than 200 agentic E2E workflows using the Playwright MCP, Playwright CLI, and agent-generated Playwright tests in test workspaces using…