patrol-qa-automation

Specialist sub-agent that composes a Patrol scenario Dart file from a confirmed list of testcase files for Flutter mobile apps. Invoked by patrol-test-creator after all testcase Dart files are saved. Handles imports, helper function orchestration, state reset between independent error paths, and end-to-end validation via Patrol MCP. DO NOT invoke for planning, triage, or testcase authoring — use patrol-test-creator instead.

Specialist sub-agent for diagnosing and fixing failing Patrol selectors in Flutter mobile apps. Use when a Patrol testcase or scenario fails because an element cannot be found, a tap does not navigate, or an assertion fails. Takes screenshot, inspects native tree, diagnoses root cause, tests fixes by running the updated file, and returns a corrected selector or Semantics change. DO NOT use for planning, triage, or writing full testcases — use patrol-test-creator instead.

Use when the user wants to create Patrol UI test files from a test case file (CSV, spreadsheet, Jira export). Handles planning, screen mapping, triage (automate vs skip), mapping table confirmation, testcase Dart authoring, and scenario composition for Flutter mobile apps. Trigger phrases: "patrol", "Patrol", "UI test", "test automation", "test script", "test scenario", "testcase", "automate", "CSV test cases", "Patrol test", "create Patrol tests", "generate patrol", "automate UI".

Autonomous Patrol debugging orchestrator for Flutter mobile apps. Use when a Patrol test exits with a failure, a testcase or scenario fails, an element is not found, a tap does not navigate, an assertion fails, or a selector breaks. Runs the failing test, captures screenshot and native tree, diagnoses root cause, applies fixes to the Dart file, re-runs to verify, and records new failure patterns in the knowledge base. Can spawn sub-agents for deep selector diagnosis, testcase rewriting, or scenario recomposition. Trigger phrases: "debug patrol", "patrol fails", "patrol error", "element not found", "test failed", "fix patrol test", "patrol assertion failed", "patrol broken".

Specialist sub-agent that writes a single atomic Patrol testcase Dart file for Flutter mobile apps. Invoked by patrol-test-creator with a confirmed test case, target screen folder, and output file name. Handles selector discovery (native tree + screen source), localization mapping, Dart authoring, syntax validation, and test execution via Patrol MCP. DO NOT invoke directly for planning or triage — use patrol-test-creator instead.

Creates Patrol UI test files (Dart) for Flutter mobile applications. Use when the user wants to write Patrol tests from a test plan, generate testcases for a screen, create end-to-end user journey scenarios, or automate mobile UI testing. Covers folder structure, selector strategies, localization patterns, and test execution.

Generate structured, exhaustive mobile UI test cases from Jira tickets, PRDs, Figma designs, or feature descriptions. Focuses exclusively on visual validation and user interactions on mobile apps (Android/iOS). Does NOT generate API, web, or backend test cases. Outputs as Jira ticket comments and consolidated CSV in /test-cases/.

Debug failing Patrol UI test files for Android/iOS mobile apps. Use when a Patrol test fails, an element is not found, a tap does not navigate, an assertion fails, or a selector breaks. Triggers: 'patrol fails', 'element not found', 'test failed', 'debug patrol', 'patrol error', 'fix patrol test', 'patrol assertion failed'.

Analyze code impact from PRs or branch diffs using tomo_search to identify affected modules and scopes in Flutter/Dart projects. Produces structured JSON output mapping impacted code to test cases that need reassessment. Uses full semantic AI matching — reads actual changed code and test case steps to determine functional impact. Output feeds into `regenerate-test-cases` skill. Trigger: impact analysis, analyze impact, which tests affected, PR impact, diff impact, what modules changed, tomo search.

Update existing mobile UI test cases when code changes are detected. Accepts git diffs, PR references, or existing test case CSVs as input. Identifies affected UI test cases and produces updated/new/removed cases. Focuses exclusively on mobile visual/interaction test cases. Does NOT generate API, web, or backend test cases. Trigger: regenerate test cases, update test cases, refresh test cases, sync test cases with code.