Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
actionbook avatar

Core Agent Browser

  • 735 installs
  • 1.4k repo stars
  • Updated May 24, 2026
  • actionbook/rust-skills

core-agent-browser is an internal agent skill that wires vercel-labs agent-browser CLI sessions into Rust agent workflows for scraping, form flows, and visual checks when actionbook MCP selectors are unavailable.

About

core-agent-browser is an internal support skill in actionbook/rust-skills marked user-invocable false and disable-model-invocation true. It orchestrates vercel-labs agent-browser CLI as the last-resort browser automation layer after rust-learner and actionbook MCP pre-computed selectors. Use core-agent-browser when actionbook lacks selectors for a target site, interactive browser testing is required, or screenshots and form filling are needed for Rust crate and documentation research. The skill defines a three-tier fetch priority: rust-learner orchestration, actionbook MCP structured manuals, then agent-browser CLI direct automation. Background agents need explicit permission to run agent-browser in project configuration. Developers indirectly benefit when rust-learner, docs-researcher, or crate-researcher skills escalate to live browser fetches for crates.io, docs.rs, or release pages missing from actionbook coverage.

  • Headless browser session control
  • DOM navigation and extraction
  • Auth and cookie handling
  • Screenshot and assertion hooks
  • Rust-native agent action bindings

Core Agent Browser by the numbers

  • 735 all-time installs (skills.sh)
  • Ranked #1,380 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
  • Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/actionbook/rust-skills --skill core-agent-browser

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs735
repo stars1.4k
Last updatedMay 24, 2026
Repositoryactionbook/rust-skills

How do Rust agents fetch pages without actionbook selectors?

Wire headless or driven browser sessions into Rust agents for scraping, form flows, visual checks, and authenticated web tasks without ad-hoc scripts.

Who is it for?

Rust-skills ecosystem agents that exhausted actionbook MCP selectors and need agent-browser CLI for live docs.rs, crates.io, or release page automation.

Skip if: Direct end-user invocation or general web scraping outside the rust-skills agent chain—core-agent-browser is internal support with user-invocable set to false.

When should I use this skill?

rust-learner, docs-researcher, or crate-researcher explicitly requires browser automation because actionbook has no pre-computed selectors or interactive testing is needed.

What you get

Headless agent-browser session output with scraped page content, screenshots, or completed form interactions for Rust research workflows.

  • Scraped web page content from headless session
  • Screenshots of target Rust documentation pages
  • Completed form interactions for authenticated fetches

By the numbers

  • Defines a 3-tier fetch priority ending with agent-browser CLI
  • Supports 3 use cases: missing selectors, interactive testing, screenshots and forms

Files

SKILL.mdMarkdownGitHub ↗

Browser Automation with agent-browser

Priority Note

For fetching Rust/crate information, use this priority order: 1. rust-learner skill - Orchestrates actionbook + browser-fetcher 2. actionbook MCP - Pre-computed selectors for known sites 3. agent-browser CLI - Direct browser automation (last resort)

Use agent-browser directly only when:

  • actionbook has no pre-computed selectors for the target site
  • You need interactive browser testing/automation
  • You need screenshots or form filling

Quick start

agent-browser open <url>        # Navigate to page
agent-browser snapshot -i       # Get interactive elements with refs
agent-browser click @e1         # Click element by ref
agent-browser fill @e2 "text"   # Fill input by ref
agent-browser close             # Close browser

Core workflow

1. Navigate: agent-browser open <url> 2. Snapshot: agent-browser snapshot -i (returns elements with refs like @e1, @e2) 3. Interact using refs from the snapshot 4. Re-snapshot after navigation or significant DOM changes

Commands

Navigation

agent-browser open <url>      # Navigate to URL
agent-browser back            # Go back
agent-browser forward         # Go forward
agent-browser reload          # Reload page
agent-browser close           # Close browser

Snapshot (page analysis)

agent-browser snapshot        # Full accessibility tree
agent-browser snapshot -i     # Interactive elements only (recommended)
agent-browser snapshot -c     # Compact output
agent-browser snapshot -d 3   # Limit depth to 3

Interactions (use @refs from snapshot)

agent-browser click @e1           # Click
agent-browser dblclick @e1        # Double-click
agent-browser fill @e2 "text"     # Clear and type
agent-browser type @e2 "text"     # Type without clearing
agent-browser press Enter         # Press key
agent-browser press Control+a     # Key combination
agent-browser hover @e1           # Hover
agent-browser check @e1           # Check checkbox
agent-browser uncheck @e1         # Uncheck checkbox
agent-browser select @e1 "value"  # Select dropdown
agent-browser scroll down 500     # Scroll page
agent-browser scrollintoview @e1  # Scroll element into view

Get information

agent-browser get text @e1        # Get element text
agent-browser get value @e1       # Get input value
agent-browser get title           # Get page title
agent-browser get url             # Get current URL

Screenshots

agent-browser screenshot          # Screenshot to stdout
agent-browser screenshot path.png # Save to file
agent-browser screenshot --full   # Full page

Wait

agent-browser wait @e1                     # Wait for element
agent-browser wait 2000                    # Wait milliseconds
agent-browser wait --text "Success"        # Wait for text
agent-browser wait --load networkidle      # Wait for network idle

Semantic locators (alternative to refs)

agent-browser find role button click --name "Submit"
agent-browser find text "Sign In" click
agent-browser find label "Email" fill "user@test.com"

Example: Form submission

agent-browser open https://example.com/form
agent-browser snapshot -i
# Output shows: textbox "Email" [ref=e1], textbox "Password" [ref=e2], button "Submit" [ref=e3]

agent-browser fill @e1 "user@example.com"
agent-browser fill @e2 "password123"
agent-browser click @e3
agent-browser wait --load networkidle
agent-browser snapshot -i  # Check result

Related skills

How it compares

core-agent-browser is the rust-skills internal browser fallback; use actionbook MCP first when pre-computed selectors exist for the target site.

FAQ

Can users invoke core-agent-browser directly?

core-agent-browser sets user-invocable false and disable-model-invocation true in rust-skills. Only rust-learner, docs-researcher, and crate-researcher should call it when browser automation is explicitly required.

When does core-agent-browser run versus actionbook MCP?

core-agent-browser runs as the third-tier fallback after rust-learner orchestration and actionbook MCP pre-computed selectors. Invoke it when actionbook lacks site coverage or interactive testing and screenshots are needed.

Which CLI does core-agent-browser use?

core-agent-browser drives the vercel-labs agent-browser CLI for headless sessions, scraping, form filling, and screenshots. Background agents must grant agent-browser execution permission in project configuration.

AI & Agent Buildingagentsautomation

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.