QA: Regression Test (Live Site) — Autonomous Agents Agentic Workflow

schema_version: 1
id: agent_qa_regression
title: "QA: Regression Test (Live Site)"
description: "Regression test affected modules on the live/deployed site after merge"
overview: |
  After a merge ships to production, this chain re-runs targeted regression checks on the modules the PR touched, plus any modules declared as adjacent in the knowledge pack. The agent loads the module knowledge packs, walks the known-good user paths, and flags any selector drift or behavioural regression it finds.

  This differs from first-pass QA in that it is not driven by acceptance criteria — the baseline is "everything the module used to do still works." Run it automatically after each merge window, or manually whenever a deploy is suspected of breaking unrelated flows.

category:
  level: pipeline
  domain: engineering

metadata:
  agent: true
  role: qa

params:
  ticket_url:
    type: string
    description: "Jira ticket URL"
  depth:
    type: string
    default: "L3"
    description: "Max test depth: L0 (smoke), L1 (structure), L2 (interaction), L3 (data)"
  hint:
    type: string
    default: ""
    description: "Optional extra instructions for the agent"
  entry_path:
    type: string
    default: "~/workspace"
    description: "Working directory for the agent"

steps:
  - type: terminal.open
    title: "Agent: QA Regression (Live)"
    cwd: "{{entry_path}}"
  - type: terminal.run
    cmd: >-
      source ~/.agent-env && claude --dangerously-skip-permissions
      "read ticket with attachments and all comments - {{ticket_url}}. if it fails, stop and report the error.
      identify which modules the ticket affects from description, PR links, and code changes.
      learn how to QA relevant modules by loading skill packs.
      run regression tests up to depth {{depth}} following the module qa.md test phases in order.
      collect screenshot evidence at phase boundaries.
      {{hint}}
      write one comment to the ticket with: modules tested, phases completed, per-phase results (test#, result, notes), new bugs found (repro steps), and overall verdict: PASS (no regressions) or FAIL."