Anahata Yam

"Yet Another Module" - The creative and experimental laboratory for multimodal agentic tools.

Yam provides specialized toolkits that extend the ASI's reach into external environments, from web automation to hardware interaction. It acts as an incubator for next-generation tools.

Chrome Automation (19 Tools)

The Chrome toolkit leverages Selenium WebDriver to give the ASI full "eyes and hands" on the web. It is engineered for secure, debug-mode interactions with an existing browser profile.

Visual Reasoning

The ASI can take screenshots of any page or specific element, enabling it to "see" UI layouts and debug front-end issues visually.

Form Metabolism

Surgical form-filling capabilities allow the model to navigate complex multi-step workflows or automate repetitive data entry tasks.

The Connection Loop
  1. Automatically detects running Chrome instances.
  2. Restarts Chrome in --remote-debugging-port mode if necessary.
  3. Maintains session persistence via a dedicated "Anahata" profile.
connect(droneId, profile, headless, ...)

Detects running browsers or launches a new Selenium WebDriver session.

navigate(droneId, url)

Directs the browser drone to the specified URL.

getPageSource(droneId)

Gets the full HTML source of the current page.

executeScript(droneId, script)

Executes arbitrary JavaScript in the current session.

getScreenshot(droneId, name)

Takes a screenshot of the current page and attaches it.

fillForm(droneId, dataMap)

Fills web forms using an intelligent multi-strategy element locator.

inspectForm(droneId)

Analyzes current page for interactive inputs and buttons.

clickElement(droneId, identifier)

Performs a stateful click via ID, Name, Link Text, or XPath.

waitForElement(droneId, cssSelector, timeout)

Waits for an element to be visible on the page.

listTabs(droneId)

Uses native Chrome DevTools Protocol to list tabs without flickering.

Firefox Automation (19 Tools)

The Firefox toolkit provides the same powerful Selenium WebDriver automation as Chrome, but specifically tailored for Mozilla Firefox and GeckoDriver.

Gecko Automation

Perfect for cross-browser testing or leveraging Firefox's unique privacy containers. Includes all 19 standard web automation tools (navigation, screenshots, form filling, script execution) implemented via GeckoDriver.

Internet Radio (3 Tools)

The Radio toolkit provides a direct audio feed for the developer's "Metabolic Flow." Integrated with the SomaFM API and high-quality streams like KEXP and FIP Paris.

Station Curation

Deeply integrated curated channels (Groove Salad, Drone Zone, Nightride FM) provided as native system instructions.

Device Targeting

Synchronous hardware line selection ensures audio is routed to the correct output device (HDMI, Headphones, System) via a custom JLayer bridge.

start(url)

Starts playing a specific internet radio station by its URL.

stop()

Stops the currently playing radio stream.

selectOutputDevice(deviceId)

Selects a specific device for radio playback.

Speech Synthesis (TTS)

Leveraging the FreeTTS engine, the Speech toolkit allows the ASI to communicate via high-salience audio alerts or full text-to-speech synthesis.

Designed for "Eyes-Free" debugging and status notifications, the toolkit provides immediate feedback on long-running tasks or critical errors using the OS-independent speak(text) tool.

speak(text)

Speaks the given text using the internal pure-Java TTS engine.

The Experimental Lab

Yam is the birthplace of all new multimodal capabilities. It serves as the incubation chamber for the next generation of Anahata tools before they are promoted to the Core or Desktop environments.