toaipa

rf-appium

Load when the user wants to write, edit, or refactor mobile automation tests (iOS/Android) — in any language (e.g. "test mobile app", "kiểm tra app iOS", "viết test Android", "automate mobile", "test ứng dụng di động"). Covers iOS/Android tests, keywords, click, input, swipe, gestures, mobile locators, capabilities, context switching. Contains mandatory rules for Sleep, locator variables, timeout variables, screenshots, coordinate clicks (prefer percentage-based), and verifying locators via Get Source or Appium Inspector. Always load together with rf-keyword-builder, rf-testcase-builder, and rf-resource-architect when writing or refactoring code. Before writing any keyword call, MUST load rf-libdoc-search to find the correct keyword name, then rf-libdoc-explain to confirm arguments — never rely on memory.

rf-browser

Load when writing any Robot Framework code using Browser Library or Playwright — including web tests, click, input, wait, assert, locators, iframes, multi-tab, file upload/download. Contains mandatory rules for Sleep, locator variables, timeout variables, screenshots, coordinate clicks, and verifying locators before use. Load immediately when Browser Library, Playwright, or web testing is mentioned, or when starting to write keywords/tests for web.

rf-results

Load when the user wants to check, analyze, or understand Robot Framework test results or failures — in any language (e.g. "check results", "xem kết quả test", "test bị fail", "tại sao test fail", "bao nhiêu test pass", "đọc output.xml"). Analyzes output.xml for pass/fail counts, detailed errors, failure pattern analysis, merging multiple outputs, and timing. After parsing, always analyze the root cause and suggest fixes. When fixing failures, also load rf-keyword-builder and the relevant library skill (rf-selenium/rf-appium/rf-requests), then MUST load rf-libdoc-search and rf-libdoc-explain before writing any fix — never rely on memory.

rf-selenium

Load when the user wants to write, edit, or refactor web browser automation tests — in any language (e.g. "test web", "kiểm tra trang web", "truy cập URL", "viết test browser", "mở trình duyệt", "open browser", "navigate to"). Covers web tests, keywords, click, input, wait, assert, forms, multi-window, iframes, JavaScript. Contains mandatory rules for Sleep, locator variables, timeout variables, screenshots, coordinate clicks, explicit waits, and verifying locators before use. Always load together with rf-keyword-builder, rf-testcase-builder, and rf-resource-architect when writing or refactoring code. Before writing any keyword call, MUST load rf-libdoc-search to find the correct keyword name, then rf-libdoc-explain to confirm arguments — never rely on memory.

rf-libdoc-explain

Load when the user needs to know how to use a specific RF keyword — arguments, syntax, options — before writing code; never guess from memory — in any language (e.g. "how to use keyword X", "keyword X dùng như thế nào", "tham số của X là gì", "cú pháp của X"). Use after rf-libdoc-search or independently when keyword name is already known. After explaining, always load the relevant library skill (rf-selenium/rf-appium/rf-requests) and rf-keyword-builder to write the code correctly.

rf-libdoc-search

Load before writing any keyword or test step when the correct keyword name is uncertain — never guess or invent names — in any language (e.g. "which keyword to use for...", "tìm keyword để...", "có keyword nào để...", "keyword nào dùng cho..."). After finding candidates, load rf-libdoc-explain for full argument details.