ego(lite) articles
Everything you need to know about browser automation
Guides, insights, comparisons, and reviews across Playwright MCP, Browser Use, Chrome DevTools MCP, coding agents, and the browsers they drive.
Guides
Why LLM agents fail at web search & how to fix it
A bad answer can begin with a missing search call, a vague query, an inaccessible page or evidence the model ignored. Inspect each stage before changing the prompt.
Oct 29, 2026 · 11 min read
Guides
Claude in Chrome: setup, permissions, and safety
A current, security-first guide to installing Claude in Chrome, choosing Manual, Auto, or Skip, connecting other Claude surfaces, and diagnosing blocked sites.
Oct 29, 2026 · 11 min read
Guides
Hermes browser extension: connect agents to your browser
Hermes can reach a local browser through several layers. This guide separates the community extension from the Hermes runtime, then compares extension, gateway, CDP, CLI, and ego(lite) by setup, sessions, and tab ownership.
Oct 29, 2026 · 10 min read
Guides
Playwright Test failure states: handling API 500s
A 500 is not one test outcome. Mock it when you own the error UI, fail immediately on a live contract, and never retry a deterministic server error until green.
Oct 27, 2026 · 9 min read
Guides
Browser automation agent skills: install and verify
Installing a browser skill gives an agent instructions. It doesn't supply a logged-in session or prove a write succeeded. We checked both questions on one owned task.
Oct 26, 2026 · 17 min read
Guides
Agent harness: what it is, how to build one, how to choose
Agent equals model plus harness. Three authorities currently disagree about where a harness ends and a runtime begins, so this starts by settling that, then builds a small one, then looks at the part almost nobody writes about: what the agent actually sees on the web.
Oct 26, 2026 · 18 min read
Guides
How to connect AI agents to your existing browser
The most-requested agent capability across forums and issue trackers, mapped honestly: four routes with their mechanisms and risks, the account and permission checks each one requires, and the two questions that pick yours.
Oct 26, 2026 · 10 min read
Guides
Cursor browser automation: MCP, fixes, and alternatives
Know the built-in Browser's real boundary before debugging around it: per-workspace isolated state, approval-gated tools, and no automatic inheritance of your personal logins. The three routes beyond it, plus fixes for common failures.
Oct 26, 2026 · 16 min read
Guides
AI scraping behind login walls: X and LinkedIn (2026)
Scraping behind a login is three routes, not one: inject a session, reuse a real browser's session, or point a no-code tool at it. For X and LinkedIn, collect only what the signed-in account is allowed to read, record evidence, and never treat a login as permission to bypass controls.
Oct 26, 2026 · 20 min read