qa-help
Skill2 starsChanged 7 days agoApache-2.0View on GitHub
Show the blueprint's command overview — every slash command, what it does, when to use it, plus natural-language alternatives. Use for "/qa-help", "help", "welche Befehle gibt es", "was kannst du", or when the user seems lost. Not for testing — route testing requests to /test-ticket.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- accessibility-testingSkill
- agentswitchSkill
- api-testingSkill
- bug-reportSkill
- bug-reproductionSkill
- demoSkill
- deploy-to-stageSkill
- exploratory-testingSkill
- failure-analysisSkill
- model-setupSkill
- onboard-productSkill
- onboard-projectSkill
- perf-analysisSkill
- qa-metricsSkill
- qa-reportSkill
- re-boardSkill
- reconSkill
- release-readinessSkill
- report-to-jiraSkill
- retest-bugSkill
- retro-learnSkill
- risk-based-testingSkill
- scaffold-suiteSkill
- security-testingSkill
- skill-reviewSkill
- test-automationSkill
- test-case-managementSkill
- test-reliabilitySkill
- test-ticketSkill
- visual-testingSkill
- webdriverio-testingSkill
- whats-newSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 70
Evidence is mandatory: every run ends with a verdict (PASS, FAIL, BLOCKED, NEEDS_REVIEW, NO_TEST) and artifacts under evidence/.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
/qa-help — what this blueprint can do
Render the overview below, then stop. Do not start any work from it — if the user picks a command, run that skill next. Answer in the user's language.
You do not need to memorize commands: plain language works. "Test SHOP-123",
"re-test the fix on WEB-456", "how is testing going" all route to the right
skill. The slash commands below are the explicit, guaranteed form. In plugin
mode they are namespaced: /qa-agent-blueprint:<name>.
Test a ticket
| Command | What it does |
/test-ticket <KEY-123> | The main entry: reads the ticket, routes to project + product, checks stage, picks the test strategy, produces evidence, reports the verdict back to JIRA. |
/retest-bug <KEY-123> | Re-plays the original repro steps of a fixed bug, then a tight regression sweep around the spot. |
/exploratory-testing | Chartered, timeboxed exploratory session via Playwright MCP — for new features or unclear behavior with no repeatable spec yet. |
/test-automation | Writes or runs deterministic specs in the project's framework (Playwright or WebdriverIO) once a flow is understood and worth repeating. |
/recon <KEY-123> | Observes the ticket's pages before any spec is written: probe run writes recon.json (real selectors, console errors, screenshots) into the run dir. |
React to a result
| Command | What it does |
/failure-analysis | Triages a FAIL before any bug is filed: real product defect, or outdated test case? Runs automatically on every FAIL. |
/bug-report | Turns a finding into a bug ticket in the project's own template. |
/report-to-jira | Posts the verdict comment + attachments + workflow transition to JIRA (composes only, while shadow mode is on). |
/deploy-to-stage | Runs the project's stage-deploy runbook when the fix isn't on stage yet. |
Show others
| Command | What it does |
/demo <KEY-123> | Presents a bug repro or finished test run to an audience: headed slow-mo replay plus a demo card, hands-on mode where you take over the browser ("lass mich probieren"), and screen recording (.webm) — only when you explicitly ask for it, never automatically. |
Set up
| Command | What it does |
/onboard-project | One-time interview per JIRA project — run this first for a new project key. |
/onboard-product | Short onboarding for a product inside an existing project. |
/re-board <KEY> | Delta-onboarding: brings an existing project up to the current blueprint schema (new fields, missing files, open TODOs) without repeating the full interview — the upgrade path after a blueprint update. |
/scaffold-suite | Creates the Playwright test project when a product has no test assets yet. |
/agentswitch <target> | Read-only check for switching harnesses (claude or opencode); never migrates or re-onboards. |
/model-setup | Configure cheap/standard/deep opencode model overrides per execution role; consumer-only, restart required. |
/skill-review | Review which skills the evidence shows as used (used/unknown with coverage, never "unused") and propose a slimmer skill profile after confirmation; integrate-only apply. |
Knowledge & meta
| Command | What it does |
/qa-report | Aggregates all evidence runs into QA metrics: pass rate, P0/P1, BLOCKED reasons, per project and product. |
/retro-learn | Captures and promotes learnings — runs at session end, or explicitly to groom. |
/whats-new | What changed in the blueprint since your projects were onboarded — and which changes call for /re-board. |
/qa-help | This overview. |
Three things worth knowing
- Shadow mode is on by default — the agent composes JIRA comments but
posts nothing until
Shadow mode: offis written in the project file. - The specialist skills (api, accessibility, security, visual, …) are pulled by qa-manager itself when the product's feature matrix calls for them — you don't invoke them directly.
- Evidence is mandatory: every run ends with a verdict (PASS, FAIL,
BLOCKED, NEEDS_REVIEW, NO_TEST) and artifacts under
evidence/.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

