qa
Skill2 starsChanged 2 days agoMITView on GitHub
QA toolkit for web portals, built from a Thai-government integration project and generalised: (1) answer-review — review the prepared answers in a meeting Q&A sheet through four lenses (business · program · data · operation) and write one Remark per row; (2) export-evidence — walk every export button (CSV / Excel / PDF / Print) on a signed-in Chrome, verify the files, and build the team's .xlsx evidence report; (3) excel-import — design, preflight, run and verify a test set for an Excel-template import (validate → submit → database); (4) ticket-report — post a QA / retest result on a Jira ticket in the team format with screenshots inline. Use WHENEVER the work is test evidence, a QA report, test case design for an import, a pre-meeting answer review, "ทดสอบ", "หลักฐานการทดสอบ", "test report", "ลงผลเทสในตั๋ว", or "review the Q&A before the meeting". Read the sub-skill's SKILL.md before acting.
In this repository23 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- new-skillSkill
- agent-kit-costSkill
- agent-kit-prompt-cacheSkill
- agent-kitSkill
- containers-podmanSkill
- containersSkill
- dependency-vuln-explainerSkill
- design-coreSkill
- design-flowSkill
- devops-changeSkill
- devops-network-diagramSkill
- devops-release-scopeSkill
- devopsSkill
- documentation-doc-styleSkill
- documentation-pdf-operationsSkill
- documentationSkill
- poc-spec-driven-builderSkill
- qa-answer-reviewSkill
- qa-excel-importSkill
- qa-export-evidenceSkill
- qa-ticket-reportSkill
- thai-mock-namingSkill
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 36
Evidence over claims: Every PASS carries a screenshot and a parsed file or a response body.
≈ “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 — sub-skills
| When the work is… | Read | Scripts |
| "are these prepared answers safe to say in the meeting", "review the Q&A tab", "ลง remark ให้ Q&A" | answer-review/SKILL.md + one lens per agent under answer-review/lenses/ | answer-review/scripts/qa.py (pull · merge · selfcheck) |
| "test the Export Excel/PDF/CSV buttons", "QA evidence workbook", "ทดสอบปุ่ม Export" | export-evidence/SKILL.md | export-evidence/scripts/collect_export.py (CDP walk) · make_qa_report.py (xlsx) |
| "test the xlsx import", "test cases for the import template", "are all dropdown values usable" | excel-import/SKILL.md + references/test-design.md | excel-import/scripts/run_web.py (login · run · report) with a project adapter.py (contract: adapter_example.py) |
| "post the test result on PROJ-123", "retest done, log it on Jira", "รายงานผลเทสลงตั๋ว" | ticket-report/SKILL.md + references/format.md | ticket-report/scripts/ticket_report.py (draft · post · --selftest) |
They chain: export-evidence and a page-timing run produce the evidence folder that
ticket-report drafts from (steps.csv / results.json → one case); excel-import's
report.xlsx and screenshots become a ticket's cases.json evidence.
Rules every sub-skill keeps
- A person signs in; the agent never types a password into a login form. Sessions are reused from a signed-in CDP Chrome or a saved Playwright storage state outside the repo.
- Read-only on production. Click only filters, query buttons, export buttons and queue downloads; never save, submit, confirm or delete on prod. Synthetic import files are validated, never submitted.
- Evidence over claims. Every PASS carries a screenshot and a parsed file or a response
body; every FAIL carries the cause (
run.json→net, the validate body, the DB row). An empty table is not a pass without a row count on screen. - Draft, show, then post. Nothing lands on a ticket or a shared sheet before the owner has
read the draft (
ticket-report.md,plan.json). - Secrets stay out of the repo and out of the output. Account files live under
~/.configor a gitignored folder; the one exception (dev/uat Test Data on a ticket) is explicit and cannot be switched on for master/prd. - Scripts run with
uv run <script>(PEP 723 inline dependencies) from the repo root; Playwright needs an installed Chrome (channel="chrome") orplaywright install chromium.
Tests
cd skills/qa && python -m pytest -q # needs: pytest openpyxl pillowDoes 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.

