← ClaudeAtlas

using-exloom-qalisted

Use when starting any QA test-case work — establishes which exloom-qa skill to use when, and the rules that hold across all of them.
Expeed-Software/exloom · ★ 0 · Testing & QA · score 72
Install: claude install-skill Expeed-Software/exloom
# Using exloom-qa Find the skill for your situation and invoke it. It pulls in the others in order. | Situation | Skill | |---|---| | Starting from a user story | `capturing-story-context` | | Context captured, need cases | `generating-test-cases` | | Reviewing, changing, or approving cases | `reviewing-test-coverage` | | Approved and ready for the board | `publishing-test-cases` | Commands: `/qa-test-init <project> <story-id>` → `/qa-test-review` → `/qa-test-publish`. ## Rules that hold everywhere **Nothing is published without an approval record.** Before approval, the only permitted tracker call is reading the work item. A `PreToolUse` hook enforces this and fails closed. **Every case must be runnable by a human** with the application open — no API calls, database queries, logs, devtools, or scripts. See `../references/human-executability.md`. **Volume matches story complexity.** Assess first, then generate within the band. See `../references/complexity-and-volume.md`. **The story is a delta, not a spec.** It will not describe navigation, prerequisites, or downstream effects. Capturing those from QA is mandatory, not optional. **Evidence over speculation.** Every case names its source. Never generate cases for capabilities with no evidence they exist — SSO, LDAP, biometrics, social login, IP restrictions and similar are not tested unless the story or context shows they apply. **Ask one question at a time.** Use `AskUserQuestion` with options and a recommendation.