BrowserSkill Browser Launcher

The Specific Problem It Solves

When using BrowserSkill (abbr. bsk) for web scraping or browser automation, a common prerequisite is having the local Edge browser open and the BrowserSkill extension connected to the local daemon. For scheduled tasks, CI/CD pipelines, or any unattended automation, this manual step of "launching the browser" becomes a blocking point. Scripts frequently fail with errors like browsers connected 0 or no browser is connected to the daemon, causing the entire automation pipeline to halt.

How the Skill Works

The core capability of the BrowserSkill Browser Launcher is to make the browser "self-available" within automation tasks. It is a pure control component, devoid of any business logic (e.g., login, data processing). Its workflow is encapsulated in the scripts/launch_browser.py script, with the following key steps:

1. Connection Probing: First, it invokes the bsk status command to detect if any browser instance is already connected to the daemon.
2. Conditional Launch: If no connection is detected, the script automatically launches the locally installed Microsoft Edge browser. Since the BrowserSkill extension is set to start with the browser, it automatically connects to the daemon.
3. Readiness Wait: After launch, the script enters a polling loop (defaulting to a max wait of 40 seconds, adjustable via the --wait parameter) to continuously check the browser connection status.
4. Result Feedback: Upon successful connection, the script exits with code 0 and prints a confirmation message. If the --start-session parameter is specified, it also outputs the new session identifier (SESSION=), allowing downstream tasks to use it directly.

It uses a clear set of exit codes to define handling for the caller:
- 0: Browser is ready; proceed with scraping or automation.
- 3: Connection timeout failure; the caller should consider a fallback strategy (e.g., use WebFetch directly) or mark the task.
- 4: Session creation failed; a retry may be attempted.

Applicable Boundaries and Considerations

This skill is a generic "launcher" with clear boundaries:
- Single Responsibility: It only handles launching the browser and ensuring the connection. It is unconcerned with what data is scraped or where it is published. Any scraping skill or automation task based on bsk should invoke it as a prerequisite step.
- Environmental Dependencies: Relies on the local Edge browser path (typically C:/Program Files (x86)/Microsoft/Edge/...) and the bsk daemon (~/.local/bin/bsk.exe). In non-standard environments, paths need confirmation.
- Statelessness and Security: The script contains no business credentials (appid/appsecret), managing only the browser lifecycle, and is safe for distribution.
- Asynchronous Connection Handling: After browser launch, the extension connection is asynchronous. The polling mechanism within launch_browser.py handles this, so callers do not need to implement their own sleep logic.

In summary, it is a foundational dependency library designed to eliminate the tedious and error-prone manual environment preparation step in automation tasks, allowing developers to focus more on the core business logic.