Triage failed Agent QA runs with evidence-backed diagnosis

Agent QA Result Triage inspects failed or interrupted Agent QA test runs and assigns them to one of eight fixed failure categories based on recorded evidence. Instead of guessing, it fetches run artifacts, step results, execution logs, and related run history, then returns a structured triage result with category, confidence level, likely ownership area, and the next evidence-backed action.

Use this skill when investigating a failed or interrupted Agent QA run, inspecting run artifacts or execution logs, comparing recent related runs for recurring failure patterns, or deciding whether a failure belongs to a test, product, hook, browser/mobile runtime, or infrastructure owner.

Do NOT use this skill to modify tests or application code; switch to agent-qa-debug-fix after triage is complete for authorized repairs. This skill does not prove root cause - it identifies the most likely failure surface based on available evidence.

You provide a failed Agent QA run identifier. The skill starts with agent_qa_get_run for run status, suite child context, steps, and attempts. It fetches evidence before deciding using tools such as agent_qa_get_run_artifact, agent_qa_get_run_steps, agent_qa_get_run_logs, and agent_qa_get_run_execution_logs. It calls agent_qa_classify_failure and uses its category as the default classification unless stronger evidence contradicts it.

You receive a concise triage result with exactly one category: timeout, appium_startup, browser_disconnect, element_not_found, assertion_failure, hook_failure, infrastructure, or unknown_failure. The output includes confidence, evidence, likely fix area, and next action.

{
  "category": "element_not_found",
  "confidence": "high",
  "evidence": ["Step 4 could not resolve the described checkout button"],
  "likely_fix_area": "test definition or changed product UI",
  "next_action": "Inspect the captured UI context, then compare the current checkout screen"
}

The skill integrates with Agent QA MCP tools (agent_qa_get_run, agent_qa_get_run_artifact, agent_qa_get_run_steps, agent_qa_get_run_logs, agent_qa_get_run_execution_logs, agent_qa_classify_failure). If MCP is unavailable, it falls back to dashboard REST APIs or Agent QA CLI output and states which evidence was unavailable.

QA engineers, test automation engineers, and DevOps teams responsible for triaging Agent QA test failures need to quickly classify failures and route them to the correct owner - test author, product team, infrastructure, or runtime environment. Classification reliability depends on retained run artifacts and logs; missing screenshots, DOM/accessibility context, device logs, or prior runs lower confidence. The skill redacts credentials, session tokens, personal data, and unrelated application content from reports.