AWARSE
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| GITHUB_PAT | No | GitHub Personal Access Token. Mentioned in the .env example; may be used for GitHub integrations. | |
| GEMINI_MODEL | No | Optional model name for Gemini, e.g. gemini-2.5-flash. | gemini-2.5-flash |
| LLM_PROVIDER | No | The LLM provider to use for healing. Options: gemini (default), claude/anthropic, or openai/compatible. | gemini |
| OPENAI_MODEL | No | Optional model name for OpenAI, e.g. gpt-4o-mini. | |
| GEMINI_API_KEY | No | Your Google Gemini API key. Required when using the default Gemini provider. | |
| OPENAI_API_KEY | No | Your OpenAI API key (or compatible provider key). Required when LLM_PROVIDER is set to openai. | |
| OPENAI_BASE_URL | No | Optional base URL for OpenAI-compatible APIs such as Ollama or vLLM, e.g. http://localhost:11434/v1. | |
| AWARSE_MOCK_HEAL | No | Set to 'true' to run in mock mode without calling an external LLM. Used in CI/CD testing. | |
| APPIUM_SERVER_URL | No | Appium server URL, e.g. http://localhost:4723. Required when AUTOMATION_FRAMEWORK=appium. | |
| AUTOMATION_FRAMEWORK | No | Automation backend to use: playwright (default), selenium, or appium. | playwright |
| TOKEN_EFFICIENT_MODE | No | When 'true' (default), uses optimized markdown element mapping to reduce LLM tokens. Set to 'false' to use raw HTML body context. | true |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| navigateC | Navigate the browser page or mobile app to a given URL. |
| click_elementB | Click/tap an element. Automatically heals the selector if it fails. |
| fill_elementB | Fill a form/input element with a value. Automatically heals the selector if it fails. |
| get_contentA | Retrieve the text content (web) or layout XML (mobile) of the current page/screen. |
| evaluate_jsA | Evaluate a JavaScript string on the current web page (Supported on Web frameworks only). |
| take_screenshotA | Take a screenshot of the current page/screen and save it locally. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| diagnose_selector_failure | Create a diagnostic assistant prompt to analyze a failed web/mobile element action. |
| generate_playwright_test | Create a prompt to generate a new Playwright test based on the active page layout. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| get_healed_logs | Get the session log of all selectors successfully healed by AWARSE. |
| get_page_dom | Get the layout snapshot (Markdown elements map for web, XML for mobile) of the active page. |
TDQS
Scored across 6 tools
Each tool has a clearly distinct purpose: navigation, clicking, filling, content retrieval, JS evaluation, and screenshots. No two tools overlap in functionality.
All tools follow a consistent verb_noun pattern (navigate, click_element, fill_element, get_content, evaluate_js, take_screenshot). The naming is uniform and predictable.
Six tools is a well-scoped set for a browser/mobile automation server, covering essential actions without unnecessary bloat or excessive granularity.
The set covers the core automation lifecycle: navigation, interaction, content extraction, JS execution, and screenshots. Missing explicit waiting or element-state checks, but these are minor gaps.