* Add auto-dismissal for alert and beforeunload dialogs
This PR adds automatic handling of JavaScript dialogs to prevent the agent from blocking indefinitely when `alert()` or `beforeunload` dialogs appear on web pages.
## Summary
Previously, when a website displayed native browser confirmation dialogs (like alerts or "Are you sure you want to leave?" prompts), agent-browser would hang waiting for manual intervention. This is a common issue since many websites use these dialogs for notifications or navigation warnings.
## Changes Made
- **Auto-dismiss functionality**: Added a background task that automatically accepts `alert` and `beforeunload` dialogs while leaving `confirm` and `prompt` dialogs for explicit handling
- **New flag**: Added `--no-auto-dialog` flag to disable automatic handling when needed
- **Environment variable**: Added `AGENT_BROWSER_NO_AUTO_DIALOG` for configuration
- **Documentation**: Updated README and docs with usage examples and configuration details
- **Tests**: Added comprehensive test coverage for flag parsing and dialog handling logic
## Implementation Details
- Only `alert` (notification-only) and `beforeunload` (navigation warning) dialogs are auto-handled for safety
- `confirm` and `prompt` dialogs still require explicit `dialog accept/dismiss` commands to ensure agents make deliberate choices for destructive actions
- The feature is enabled by default since these dialog types rarely require user decision-making
- Uses Chrome DevTools Protocol's `Page.handleJavaScriptDialog` for reliable dialog dismissal
Fixes#1070
* Log dialog type and message before auto-dismissal
Without this, auto-dismissed alert/beforeunload dialogs are silently
swallowed and the agent has no way to see what the dialog said. Adding
an eprintln before the CDP call makes the dismissal visible in stderr
for debugging.
* Log dialog dismissal errors instead of silently discarding them
- Remove premature "accepted" from log message since it fires before
the CDP command executes
- Replace `let _ =` with `if let Err(e)` to log failures when
Page.handleJavaScriptDialog fails
- Apply rustfmt to auto-dialog tests
---------
Co-authored-by: ctate <366502+ctate@users.noreply.github.com>
* dashboard
* fix: re-apply download behavior on recording context (#1019)
* fix: re-apply download behavior on recording context
record start creates a new browser context via Target.createBrowserContext.
Browser.setDownloadBehavior called at launch only applies to the default
context, so downloads in the recording context are silently dropped.
Fix:
1. Store download_path on BrowserManager (from LaunchOptions)
2. After creating the recording context, call Browser.setDownloadBehavior
with the new browserContextId
This ensures downloads work during recording.
Fixes#1018
* fix: add download_path to third BrowserManager constructor (auto_connect_cdp)
* fix: reap zombie Chrome process and fast-detect crash for auto-restart (#1023)
When Chrome crashes (e.g. SIGTRAP from CHECK() assertion), the daemon
now:
1. Reaps the zombie immediately via a SIGCHLD handler in the event loop
that calls waitpid(-1, WNOHANG)
2. Detects the crash instantly on the next command via a non-blocking
try_wait() check (has_process_exited), avoiding the 3-second CDP
timeout that is_connection_alive() would incur
3. Auto-relaunches Chrome transparently for the caller
Fixes#1017
Co-authored-by: ctate <366502+ctate@users.noreply.github.com>
* fix: route keyboard type through text input (#1014)
* fix: handle --clear flag in console command (#1015)
The console and errors commands parsed --clear from CLI args but the
action handlers silently ignored the flag. The handlers did not accept
the cmd parameter so they had no way to read the clear field.
Changes:
- Add clear_console() method to EventTracker in network.rs
- Update handle_console to accept cmd, read the clear field, and clear
the buffer when --clear is passed (returns {cleared: true})
- Update call site in execute_command to pass cmd
Co-authored-by: xuyongliang <yongliang.xyl@alibaba-inc.com>
* chore: patch release - ### Bug Fixes
- **Re-apply download behavior on r... (#1025)
* Add runtime stream enable/disable/status commands (#951)
* Add runtime stream management commands
* Run rustfmt and satisfy clippy
* Fix stream disable cleanup semantics
* Format stream disable regression tests
* fix: retain radio/checkbox elements in compact snapshot tree (#1008)
compact_tree() checked for "[ref=" to identify lines worth keeping, but
radio and checkbox elements render as e.g. [checked=false, ref=e1] where
the "[" opens before "checked=", not "ref=". Dropping the leading bracket
so the check is just "ref=" fixes the match for all elements with refs.
Fixes#1006
Co-authored-by: ctate <366502+ctate@users.noreply.github.com>
* chore: version packages (#1027)
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* fixes
* dashboard
* fixes
* remove observe
* fmt
* fixes
* fixes
* jotai
* fmt
* upload dashboard
---------
Co-authored-by: Stefan Smiljkovic <stefan@vanila.io>
Co-authored-by: ctate <366502+ctate@users.noreply.github.com>
Co-authored-by: zhanba <c5e1856@gmail.com>
Co-authored-by: xuyongliang <478439790@qq.com>
Co-authored-by: xuyongliang <yongliang.xyl@alibaba-inc.com>
Co-authored-by: Thomas Kosiewski <thoma471@googlemail.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
The console and errors commands parsed --clear from CLI args but the
action handlers silently ignored the flag. The handlers did not accept
the cmd parameter so they had no way to read the clear field.
Changes:
- Add clear_console() method to EventTracker in network.rs
- Update handle_console to accept cmd, read the clear field, and clear
the buffer when --clear is passed (returns {cleared: true})
- Update call site in execute_command to pass cmd
Co-authored-by: xuyongliang <yongliang.xyl@alibaba-inc.com>
* fix: handle proxy authentication via CDP Fetch.authRequired
Chrome's --proxy-server flag does not support credentials embedded in
the URL. When a proxy requires authentication, Chrome receives a 407
from the proxy but has no way to respond with credentials, resulting
in net::ERR_INVALID_AUTH_CREDENTIALS.
Fix by:
1. Parsing credentials from the proxy URL (already done by parse_proxy)
2. Storing them in DaemonState.proxy_credentials
3. Enabling Fetch.enable with handleAuthRequests: true
4. Responding to Fetch.authRequired events with Fetch.continueWithAuth
5. Passing only the server URL (without credentials) to --proxy-server
6. Forwarding credentials to the daemon via dedicated env vars
Also adds fallback to standard proxy env vars (HTTP_PROXY, HTTPS_PROXY,
ALL_PROXY, NO_PROXY) when AGENT_BROWSER_PROXY is not set.
Fixes#990
* refactor: use typed struct for parse_proxy, fix double Fetch.enable and username-only auth
- Replace serde_json::Value return from parse_proxy with a typed ParsedProxy struct
- Fix double Fetch.enable call when both proxy auth and domain filter are active
(the second call could overwrite handleAuthRequests from the first)
- Allow username-only proxy auth (some proxies don't require a password)
- Handle empty username/password in parse_proxy as None instead of Some("")
- Use install_domain_filter_fetch in auto_launch for consistency
- Update unit tests to use typed struct fields
---------
Co-authored-by: ctate <366502+ctate@users.noreply.github.com>
Fixes#992
When a JavaScript dialog (alert/confirm/prompt) blocks the page, agents
had no way to detect it — all commands just timed out with generic errors.
- Add `dialog status` command to check for pending dialogs
- Track dialog state via CDP Page.javascriptDialogOpening/Closed events
- Auto-inject `warning` field into all command responses when a dialog is
pending, so agents can distinguish dialog-blocked timeouts from other issues
- Document dialog commands in SKILL.md (was missing entirely), README.md,
docs site, and --help output
Co-authored-by: ctate <366502+ctate@users.noreply.github.com>
* feat: add network request detail and filtering for request tracking
- Add `network request <requestId>` command to view full request/response
details including response body via CDP Network.getResponseBody
- Add --type, --method, --status filter flags to `network requests`
- --type: comma-separated resource types (xhr,fetch,document)
- --method: filter by HTTP method
- --status: supports exact (200), class (2xx), range (400-499)
- Extend TrackedRequest with request_id, post_data, status,
response_headers, mime_type fields
- Update Network.responseReceived handler to also populate
tracked_requests (previously only updated HAR entries)
- Add tests for parse commands and matches_status_filter
- Update README, SKILL.md, docs, and help text
Closes#932
* fix: show request ID and status in network requests output
* fix: add ref for cursor-interactive content roles
* fix: format
* feat: always include cursor-interactive elements in snapshot, -C is deprecated
* feat: process StaticText aggregation and deduplication
* update test
* clean up
* fix: escape text of elements in snapshot
* fix: redundant slicing
* fix: cargo fmt
* feat: deduplicate redundant StaticText
---------
Co-authored-by: 羲洋 <lipengyang.lpy@alibaba-inc.com>
* fix: prevent state commands from starting daemon without session_name
(#677)
State management commands (state_list, state_show, state_clear,
state_clean, state_rename) are pure file operations that don't need a
running daemon. Previously, these commands would trigger daemon
startup
via ensure_daemon(), and if AGENT_BROWSER_SESSION_NAME was exported
after the first command (e.g. `state clear --all`), the daemon would
start without session_name. Subsequent open/close commands would
reuse
that daemon, causing close to skip state persistence entirely.
Fix: execute state management commands locally in the CLI process
before
ensure_daemon() is called. This is done via a new
dispatch_state_command() function in state.rs that centralizes the
command routing, used by both the CLI (local path) and the daemon
(batch/IPC path).
Also:
- Add OutputOptions::from_flags() helper to deduplicate construction
- Add unit tests for dispatch_state_command routing and error
handling
* style: fix fmt and clippy warnings
- Remove redundant closure in dispatch_state_command (clippy::redundant_closure)
- Remove needless borrow in run_batch (clippy::needless_borrow)
- Fix trailing blank lines (rustfmt)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: hyunjinee <leehj0110@kakao.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Navigate with load, then wait for username/password/submit selectors using the default action timeout. This avoids networkidle hangs on pages with continuous background requests.
Adds a new `agent-browser upgrade` command that automatically detects the installation method (npm, Homebrew, or Cargo) and runs the appropriate update command.
**Changes:**
- Added new `upgrade.rs` module with upgrade logic
- Updated `main.rs` to handle the `upgrade` command
- Added upgrade help text in `output.rs`
- Updated README.md and documentation with upgrade instructions
- Updated SKILL.md to mention the upgrade command
**Implementation details:**
- Fetches latest version from npm registry to show version diff
- Auto-detects installation method by checking Homebrew, Cargo paths, and npm global packages
- Provides fallback instructions if installation method cannot be determined
- Uses existing color module for consistent styled output
- Gracefully handles network failures and continues with upgrade
Fixes#895
* refactor: make --full/-f a command-level flag instead of global
Move --full/-f from global flags (parsed in flags.rs) to command-level
parsing in commands.rs, scoped to the three commands that actually use
it: `screenshot`, `diff screenshot`, and `diff url`.
This frees up `-f` for other commands (e.g. `--follow` on
`console`/`errors`, see #867) and better reflects that full-page
capture is not a global concern.
Changes:
- Remove `full` from Flags struct, Config struct, and global flag parsing
- Remove `--full`/`-f` from clean_args global boolean flags list
- Parse `--full`/`-f` inline in `screenshot` command handler
- Accept `-f` shorthand in `diff screenshot` and `diff url` (previously
only `--full` was accepted at command level)
- Remove fallback from global `flags.full` in diff subcommands
- Update tests to pass --full as a command argument rather than a global flag
Fixes#876
* fix: remove stale AGENT_BROWSER_FULL env var from help and add -f shorthand tests
- Remove AGENT_BROWSER_FULL from help text in output.rs since the env
var is no longer read after moving --full to command-level parsing
- Add test_screenshot_full_page_shorthand to verify screenshot -f works
- Add test_diff_screenshot_command_full_flag_shorthand to verify
diff screenshot -f works
---------
Co-authored-by: ctate <366502+ctate@users.noreply.github.com>
Expose HAR recording as a CLI subcommand under the existing `network`
command so users can capture and export network traffic without a
separate tool or opening the browser twice.
- Parse `network har start` and `network har stop [path]` in commands.rs
- Enrich HarEntry with request/response headers, timestamps, status text,
resource type, HTTP version, and body sizes from CDP events
- Produce HAR 1.2 output with creator/browser metadata, query strings,
and proper header arrays compatible with Chrome DevTools HAR viewer
- Auto-generate output path under ~/.agent-browser/tmp/har/ when omitted
- Add har_stop to skip_launch list so export works without a live browser
- Update help text, README, docs site, SKILL.md, and security policy docs
- Add unit tests for parsing, HAR entry serialization, and stop behavior
Add `batch` command that reads a JSON array of commands from stdin
and executes them sequentially against the daemon. This avoids
per-command process startup overhead when AI agents run multi-step
browser workflows.
Supports --bail to stop on first error (default: continue all)
and --json for structured output as an array of results.
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add screenshot output config, clipboard CLI commands, and fix wait --text native path
## Summary
- Add `--screenshot-dir`, `--screenshot-quality`, and `--screenshot-format` CLI flags (with corresponding `AGENT_BROWSER_SCREENSHOT_DIR`, `AGENT_BROWSER_SCREENSHOT_QUALITY`, `AGENT_BROWSER_SCREENSHOT_FORMAT` env vars) so users can configure where and how screenshots are saved without specifying a full path every time
- Add `clipboard read`, `clipboard write <text>`, `clipboard copy`, and `clipboard paste` CLI commands, exposing the existing protocol-level clipboard handlers that were previously only accessible via JSON-RPC
- Fix `wait --text` in native mode: the CLI was emitting `selector: "text=..."` (a Playwright-style locator) which native's `querySelector` can't handle. Now emits a `text` field that correctly hits the native `wait_for_text` polling path
- Add native clipboard `copy` and `paste` support via CDP `Input.dispatchKeyEvent`, and a `write` operation to the Node.js handler
* fix: resolve CI failures in Rust formatting and TypeScript typecheck
Use string-based page.evaluate for clipboard writeText to avoid
referencing `navigator` in Node.js compilation context. Run cargo fmt
to fix formatting in commands.rs and screenshot.rs.
* fix: clipboard write captures full multi-word text
Use rest[1..].join(" ") instead of rest.get(1) so unquoted multi-word
input like `clipboard write hello world` sends the full string rather
than silently dropping everything after the first word.
* improvements
* fixes
* improvements
* improvements
* feat: Add browserless as a hosted option + boolean env-parsing utility
* Add ensureDomainFilter, sanitizeExistingPage and move parseBooleanParam
* Add docs in relevant places, fix utils, rename of API env var
* Update readme
* Fix env variable name in readme
* Cleanup session stop urls when errors happen
* Fix browserlessStopUrl not being assigned in happy path
Bug 1: `cookies clear` printed 'Request log cleared' instead of 'Cookies cleared'
because the output handler matched the generic `{ cleared: true }` response shape
without checking the action context. Now uses the `action` parameter to distinguish
`cookies_clear` from `requests --clear`.
Bug 2: `tab close` printed 'Browser closed' instead of 'Tab closed' because the
output handler matched the generic `{ closed: ... }` response shape without checking
the action context. Now uses the `action` parameter to distinguish `tab_close` from
`close` (full browser close).
Closes#556
* add security hardening features
- Add authentication vault (`auth save/login/list/show/delete`) so credentials are stored locally and never exposed to the LLM (fixes Snyk W007)
- Add `--content-boundaries` flag to wrap page-sourced output in structural markers, helping LLMs distinguish tool output from untrusted page content (fixes Snyk W011)
- Add `--allowed-domains` flag to restrict browser navigation to trusted domains
- Add `--action-policy` for static allow/deny gating of action categories, with opt-in `--confirm-actions`/`--confirm-interactive` for orchestrator or human-in-the-loop confirmation
- Add `--max-output` flag to truncate large page outputs, preventing context flooding
- New docs page at /security, updated README, SKILL.md, CLI help text, and templates
* fixes
* fixes
* fixes
* fixes
* fixes
* fixes
* fixes
* docs
* add --selector flag to scroll command
The `scroll` command uses `window.scrollBy()`, which has no effect on apps
that use custom scrollable containers (e.g. a nested div with overflow-y: auto).
The backend `handleScroll` already supports a `selector` parameter, but the CLI
never exposed it. This adds `-s` / `--selector` to the `scroll` command so users
can target a specific scrollable element:
agent-browser scroll down 500 --selector "div.scroll-container"
Also fixes the backend to apply `direction`/`amount` when a selector is present
(previously those fields were only used in the no-selector branch).
Closes#501
* fixes
* feat: add --download-path option
Adds a `--download-path` flag (and `AGENT_BROWSER_DOWNLOAD_PATH` env / `downloadPath` config key) to set a default download directory for browser downloads.
Without this, Playwright stores downloads in a temp directory that is deleted when the browser closes. The new option passes through to Playwright's `downloadsPath` on `launch()` and `launchPersistentContext()`.
Fixes#507
* improvements
* fixes
* fixes
Adds `keyboard type` and `keyboard insertText` subcommands that
operate on the currently focused element without requiring a selector.
Essential for contenteditable editors (Lexical, ProseMirror, CodeMirror,
Monaco) where `type <selector>` doesn't trigger the editor's internal
event pipeline (beforeinput/DOM mutation).
- `keyboard type <text>` — page.keyboard.type() with real keystrokes
- `keyboard insertText <text>` — page.keyboard.insertText()
Note: `keyboard press` intentionally omitted — the existing top-level
`press` command already operates on current focus.
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Fixes#519. Playwright defaults `colorScheme` to `light` on all new contexts, overriding the browser/OS dark mode setting. This is especially disruptive in CDP mode, where every reconnection resets the scheme. The `set media dark` command also didn't persist its choice to new tabs or pages.
- Add `--color-scheme <dark|light|no-preference>` flag, config key (`colorScheme`), and env var (`AGENT_BROWSER_COLOR_SCHEME`)
- Store the preference in `BrowserManager` and automatically apply it to all new contexts (via Playwright's context option) and all new pages (via `page.emulateMedia` in `setupPageTracking`)
- `set media dark/light` now also persists its choice for subsequent pages and tabs
* feat: Enable capture of profiling data
Adding a new set of commands:
```
agent-browser profiler start
agent-browser profiler stop trace.json
```
With this, agents can start a profiling trace, perform a set of actions, and then extract the profiling data for analysis.
**Note:** I was originally going to call it `agent-browser profile` but I realized that might cause confusion with the `--profile` flag
CDP supports a couple commands for starting/stopping a trace.
When a trace is running, it emits events that need to be picked up.
We store these locally in the daemon until the trace is completed.
When the final event is received, we dump all of them into an output file.
That file can be loaded directly into chrome devtools or another analysis tool to visualize what happened during the agentic run.
Added some basic rust tests for parsing the commands (since they have some optional / required args)
TS daemon adds ~6 tests to make sure the profiling lifecycle (including saving the output file) works as intended
* add docs
* fixes
* fixes
---------
Co-authored-by: Chris Tate <chris@ctate.dev>
## Summary
- Fix `allowFileAccess` being silently stripped from launch commands by adding it to the Zod schema in `protocol.ts` (the `--allow-file-access` CLI flag was not reaching the browser)
- Fix `trace stop` requiring a path argument despite help text documenting it as optional -- now works with or without a path
- Fix `addscript`/`addstyle` silently succeeding when neither `content` nor `url` is provided -- now returns a validation error
- Replace hardcoded ANSI escape code with `color::error_indicator()` in `main.rs` to respect `NO_COLOR`
- Fix double-parse pattern and add descriptive expect messages in `commands.rs`
- Fix incomplete string escaping in `snapshot.ts` `buildSelector` (use `JSON.stringify` instead of manual quote escaping)
- Simplify redundant ternary in `snapshot.ts` cursor-interactive role assignment
- Sync docs changelog with CHANGELOG.md (v0.8.1 through v0.10.0)
* docs: fix 6 documentation issues (#303, #245, #186, #134, #61, #73)
Addresses six open documentation issues in a single pass:
- **#303** -- Add `npx agent-browser` usage across README, SKILL.md, docs site, and `--help` output for zero-install experience. Global install is recommended as the fastest path (native Rust CLI vs Node.js indirection with npx).
- **#245** -- Document Claude Code skill installation with `npx skills add vercel-labs/agent-browser`
- **#186** -- Split installation instructions into Global (recommended), Quick Start (npx), and Project (local dependency) sections with clear guidance on when to use each
- **#134** -- Add "Why agent-browser over playwright-mcp?" comparison table to README covering output format, element selection, protocol, sessions, performance, mobile, cloud, and streaming
- **#61** -- Add "Timeouts and Slow Pages" section to SKILL.md documenting the 60s default timeout, all `wait` variants, and guidance for slow websites
- **#73** -- Replace stale `cp node_modules/...` advice with `npx skills add`, add warning against copying SKILL.md manually, add "Session Management and Cleanup" section to SKILL.md
* remove section
* fix doc
Rebased and fixed implementation of PR #184 features on current main:
Session persistence:
- --session-name flag and AGENT_BROWSER_SESSION_NAME env var auto-save/restore
cookies and localStorage across browser restarts
- State files stored in ~/.agent-browser/sessions/ with owner-only permissions
- AES-256-GCM encryption via AGENT_BROWSER_ENCRYPTION_KEY env var
- Auto-expiration of old state files (AGENT_BROWSER_STATE_EXPIRE_DAYS, default 30)
State management commands:
- state list: list saved state files with metadata
- state show <file>: display state summary (cookies, origins, domains)
- state rename <old> <new>: rename state files
- state clear [name] [--all]: clear saved states
- state clean --older-than <days>: delete expired states
New --new-tab flag for click command:
- Opens link href in a new tab instead of navigating the current tab
Security hardening:
- Session name validation prevents path traversal (CLI + daemon)
- safeHeaderMerge prevents prototype pollution in header merging
- WebSocket stream server binds to 127.0.0.1 only
- State files written with 0o600 permissions
Fixes applied over the original PR:
- Use color.rs module instead of hardcoded ANSI escape codes
- Align CLI output field names with daemon response format
- Add CLI-level --session-name validation (not just daemon-side)
- Avoid adding "DOM" to tsconfig.json lib (use proper typing in evaluate)
- Keep version at 0.9.3 (matches current main)
- Centralize session name validation in daemon.ts helper
- Update all documentation (README, SKILL.md, docs site, --help output)
Co-authored-by: Chris Tate <chris@ctate.dev>
* cargo fmt
* fix: remove redundant `use libc` import (clippy::single_component_path_imports)
* fix: use `.first()` instead of `.get(0)` (clippy::get_first)
* fix: use `.copied()` instead of `.map(|s| *s)` (clippy::map_clone)
* fix: allow too_many_arguments on ensure_daemon (clippy::too_many_arguments)
* fix: use `then_some` instead of `then` with closure (clippy::unnecessary_lazy_evaluations)
* fix: use pattern match instead of redundant guard (clippy::redundant_guards)
* fix: use pattern match instead of redundant guard in commands.rs (clippy::redundant_guards)
* fix: use `contains()` instead of `iter().any()` for simple equality (clippy::manual_contains)
* Add changeset
* feat: add --allow-file-access flag for file:// URL support
Adds the ability to open and interact with local files using file:// URLs.
This enables use cases like viewing local PDFs, testing local HTML files,
and allowing JavaScript to access other local files via XHR.
The flag adds Chromium's --allow-file-access-from-files and --allow-file-access
launch arguments. Only supported in Chromium browsers.
Fixes#345
* fix: add cli_allow_file_access tracking to prevent spurious warning
When --allow-file-access is set via AGENT_BROWSER_ALLOW_FILE_ACCESS env var
(not CLI), don't warn about the flag being ignored when daemon is already running.
* fix: only warn about ignored flags when explicitly passed via CLI
The warning about launch-time options being ignored (when daemon is
already running) was incorrectly shown when options were set via
environment variables like AGENT_BROWSER_EXECUTABLE_PATH, even when
no CLI flag was passed.
Now the warning only appears when flags are explicitly passed on the
command line, not when values come solely from environment variables.
Fixes#372
* feat: add cursor-interactive element detection in snapshots
Add -C/--cursor flag to snapshot command that detects clickable elements
that don't have proper ARIA roles but are interactive based on:
- cursor: pointer CSS style
- onclick attribute/handler
- tabindex attribute
This helps with modern web apps that use custom divs/spans as buttons.
Fixes#366
* fix: add cursor option to getSnapshot type signature
* Add base64 input for eval command
Adds -b/--base64 flag to decode script from base64, avoiding shell escaping issues for AI agents.
* Document base64 eval in SKILL.md
