core-agent-browser
actionbook/rust-skills
Internal browser automation CLI for agent workflows requiring interactive page testing and form filling.
What is core-agent-browser?
core-agent-browser is an internal support skill providing CLI-based browser automation for rust-learner, docs-researcher, and crate-researcher agents. Use it only when actionbook selectors are unavailable or interactive testing, screenshots, and form filling are explicitly required.
- Navigate pages and manage browser state (open, back, forward, reload, close)
- Capture page snapshots with interactive element references for reliable interaction
- Click, fill, type, hover, check, select, and scroll elements using snapshot-derived refs
- Press keys and key combinations, wait for elements or network conditions
- Take full-page and targeted screenshots
- Retrieve element text, values, page titles, and URLs
How to install core-agent-browser
npx skills add https://github.com/actionbook/rust-skills --skill core-agent-browser- agent-browser CLI installed and available in PATH
How to use core-agent-browser
- 1.Run `agent-browser open <url>` to navigate to the target page
- 2.Execute `agent-browser snapshot -i` to retrieve interactive elements with reference IDs (e.g., @e1, @e2)
- 3.Interact with elements using their refs: `agent-browser click @e1`, `agent-browser fill @e2 "text"`
- 4.Use `agent-browser wait` to pause for elements or network conditions if needed
- 5.Re-snapshot after navigation or DOM changes to get updated element references
- 6.Close the browser with `agent-browser close` when done
Use cases
- Testing form submission workflows when actionbook has no pre-computed selectors
- Automating interactive browser tasks like filling multi-step forms or clicking dynamic elements
- Capturing screenshots of rendered pages for verification or debugging
- Waiting for network idle or specific elements to appear before proceeding
- Interacting with sites where static selector extraction is insufficient
- Rust learning agents (rust-learner)
- Documentation research agents (docs-researcher)
- Crate information research agents (crate-researcher)
- Agent developers needing fallback browser automation when MCP selectors unavailable
core-agent-browser FAQ
Use actionbook first for pre-computed selectors on known sites, then rust-learner for orchestrated Rust/crate lookups. Use core-agent-browser only as a last resort when actionbook has no selectors and you need direct browser automation, screenshots, or form filling.
The snapshot returns element references like @e1, @e2, etc. Use these refs in commands: `agent-browser click @e1`, `agent-browser fill @e2 "value"`, `agent-browser select @e1 "option"`.
Re-run `agent-browser snapshot -i` to get updated element references, since the DOM may have changed and old refs may no longer be valid.
Yes. You can use semantic locators like `agent-browser find role button click --name "Submit"` or `agent-browser find text "Sign In" click` as an alternative to snapshot refs.
No. core-agent-browser is an internal support skill for agent workflows only and cannot be directly invoked by users.
Full instructions (SKILL.md)
Source of truth, from actionbook/rust-skills.
name: core-agent-browser description: "Internal support skill for agent-browser CLI workflows used by rust-learner, docs-researcher, and crate-researcher. Use only when browser automation is explicitly required." user-invocable: false disable-model-invocation: true
Browser Automation with agent-browser
Priority Note
For fetching Rust/crate information, use this priority order:
- rust-learner skill - Orchestrates actionbook + browser-fetcher
- actionbook MCP - Pre-computed selectors for known sites
- 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
- Navigate:
agent-browser open <url> - Snapshot:
agent-browser snapshot -i(returns elements with refs like@e1,@e2) - Interact using refs from the snapshot
- 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
More from actionbook/rust-skills and the wider catalog.
core-dynamic-skills
Internal orchestrator for generating and managing Rust crate-specific skills from project dependencies.
core-fix-skill-docs
Internal maintenance tool for checking and fixing Rust skill documentation references.
domain-cli
Build ergonomic CLI tools with type-safe argument parsing, config layering, and user feedback.
domain-cloud-native
Build cloud-native Rust apps with stateless design, graceful shutdown, and observability patterns.
domain-embedded
Embedded Rust development constraints: no_std, no heap, interrupt safety, and hardware ownership patterns.
domain-fintech
Fintech domain constraints and patterns for Rust: precision, immutability, and audit trails.