Skip to content
Back to skills

Qa Backend

ASecurity

QA backend logic: design and run unit, integration and edge-case tests for services, jobs, database code and business rules. Use when asked to test a backend change, find bugs in a service, or raise test coverage where it matters.

  • 5 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 11, 2026
ai-agentsdatabasebackend

Security analysis

A100/100

Scanned November 11, 2026

npx -y skills add thameema/memnos --skill qa-backend --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Qa Backend?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Qa Backend
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/thameema-qa-backend/badge)](https://www.skillsdirectory.com/skills/thameema-qa-backend)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: qa-backend
description: QA backend logic: design and run unit, integration and edge-case tests for services, jobs, database code and business rules. Use when asked to test a backend change, find bugs in a service, or raise test coverage where it matters.
---

# Backend QA

Aim tests at the rules that cost money or data when they are wrong, not at coverage numbers.

1. **Read the change and its callers.** List the rules it must enforce and the invariants it must not break.
2. **Choose the level.** Pure logic gets unit tests. Anything involving the database, queue or transaction gets an integration test against the real engine the project uses, not a mock of it.
3. **Cases to cover for each rule:** the normal case, boundaries (0, 1, max, empty, null), duplicates and retries (is it idempotent?), concurrent callers, ordering, time and timezone edges, and partial failure (the second write fails after the first succeeded).
4. **Money, auth and deletion paths** get the most cases: wrong user, wrong tenant, expired or revoked credentials, double charge, undo.
5. **Write the failing test first** when you suspect a bug, watch it fail for the right reason, then report it or fix it as asked.
6. **Run the project's existing suite** with its own command and report real output. Do not mark anything passing that you did not run.

## Report

Tests added (file and what each proves), bugs found with the failing input, and anything you could not test (no test database, external service, flaky timing). Keep tests deterministic: no sleeps, no network, no shared state between tests.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in ai-agents →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.