Add agent-browser skill installation to configuration section

The agent-browser CLI ships its own marketplace with a first-party skill
that Claude needs to understand the ref lifecycle and command API. Placed
the install instructions in Configuration > Plugins and Skills where
other marketplace setup lives, with a cross-reference from the Usage >
Web Browsing section.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: dguido

README.md

@@ -336,6 +336,15 @@ claude plugin marketplace add trailofbits/skills-curated
 
 For external marketplaces (Anthropic official, superpowers, compound-engineering, etc.), see [skills-curated](https://github.com/trailofbits/skills-curated) -- it maintains the approved list and install scripts.
 
+#### agent-browser skill
+
+The `agent-browser` CLI (installed in [Prerequisites](#tools)) ships its own marketplace with a first-party skill that teaches Claude the snapshot/ref workflow, command syntax, session management, authentication flows, video recording, and proxy support (~2,000 lines of reference material plus reusable shell templates). agent-browser is new enough that it's not in the model's pretraining data -- without this skill, Claude won't know the ref lifecycle or command API.
+
+```
+/plugin marketplace add vercel-labs/agent-browser
+/plugin install agent-browser@agent-browser
+```
+
 #### Publishing skills
 
 Where to publish depends on the audience:
@@ -486,6 +495,8 @@ agent-browser fill @e2 "text"   # Fill input
 agent-browser screenshot        # Capture screenshot
 ```
 
+You need to install the first-party skill for Claude to use agent-browser effectively -- see [agent-browser skill](#agent-browser-skill) in Configuration.
+
 #### Claude in Chrome (MCP)
 
 Browser automation via the [Claude in Chrome](https://chromewebstore.google.com/detail/claude/fcoeoabgfenejglbffodgkkbkcdhcgfn) extension. Operates inside your actual Chrome browser, so it has access to your existing login sessions, cookies, and extensions. This is the only option that can interact with authenticated pages (Gmail, Google Docs, Jira, internal tools) without re-authenticating. The tradeoff is that it uses screenshots and accessibility trees for page understanding, which consumes more context than agent-browser's ref system.