Page 1
How browser actions work
Think in tabs, snapshots, refs, and stable targeting instead of brittle selector guessing.
Browser profiles explained
Choose between isolated, existing-session, and remote attachment intentionally.
OpenClaw section
Browser guide
Use this section when the work depends on real pages, login state, responsive QA, file flows, and browser-specific recovery rather than generic web advice.
Core idea
Strong browser work starts with the right profile, the right evidence, and the right troubleshooting split.
Core concepts
Build the right browser model first
Page 2
Page 3
State, cookies, storage, and emulation
Reproduce environment-sensitive behavior deliberately instead of calling it random.
Practical workflows
Move from browsing to operator workflows
Page 5
Mobile, responsive, and render QA
Capture repeatable evidence for layout and device-sensitive issues.
Page 6
Screenshots, PDFs, and output capture
Choose the right artifact for review, export, and debugging.
Troubleshooting and routing
Separate browser health from page reachability and session mistakes
Page 8
Session, profile, and login-state troubleshooting
Fix wrong-context browser behavior without faking deeper breakage.
Page 9
Local preview, remote routing, and node-host control
Debug preview access and machine-split browser topology cleanly.