Index of every page on this site: llms.txt
All skills
Run it yourself

Use when asked for test cases, QA coverage, or how to test a flow by hand. Not for writing the test code or a security audit.

  • 1 file
  • 3.7 KB
  • Updated 3 weeks ago
  • GitHub
Use this Skill: https://skilld.dev/gh/avik-creator/agent-skills/test-cases

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈34 tokens for metadata: the name and description. ≈904 when used: this file.

Description uses 1.7% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈138 characters, or 1.7%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

Test cases

Scope

Do not edit code. coding and testing are the skills that write code. Do not write anything until asked. Cover a security requirement or access rule the source already states. Do not invent controls.

Write the smallest set that would catch a real failure. Skip a case that does not change the outcome.

Before writing

Ask every blocking question in one turn. Never invent a fact. Put an unknown under Open questions and name what it blocks. Replies are 1–2 lines. Write a paragraph only when the user asks.

Ask what would change the expected behavior, data, environment, limits, or permissions. An unknown expected result is not a case. Do not write a result of "TBD" or "works correctly."

What to read

  • A PRD or spec: requirements, acceptance criteria, rules, states, constraints, and non-goals.
  • Code or a diff: the changed behavior, its contract, its callers, and the conditions that changed. Stop at the affected path.
  • A bug report: the shortest reproduction, expected versus actual behavior, and the environment when it changes the result.

If the repository already tests a behavior, say so. Do not ask testing to add a second copy, and do not repeat it as a manual case.

Which cases

Cover the relevant happy path, failure, boundary, state change, permission, and regression. Use one representative value per equivalence class. A boundary is the edge: the last allowed value and the first rejected value. Boundaries are not only numbers: a daylight-saving change, Feb 29, month end, a list at page size N and N+1, and a string that is only spaces.

Test one invalid condition at a time. Combine inputs only when they change the outcome together. One behavior is one case, even when the spec and the code both mention it.

Every rule the source states gets a case, or a line under the summary that says it is not covered and why. Skip a behavior the spec lists as a non-goal. A missing behavior that loses data, money, access, or the main flow gets a case, or an open question that says why it cannot be tested yet.

A bug gets one case: it fails now and passes once the expected behavior is true. A flow gets the valid path plus the invalid steps a person can actually try: opening a later step by its URL, going back, refreshing, submitting twice, two people doing the same step at once, and what is left after an error.

These rules stay inside the skill. Do not print them as labels on a case.

The list

Each case has these parts, in this order, in plain language a person can follow without the codebase:

  • Case: a short name of the behavior.
  • Prerequisites: only when the case needs a starting state. Leave this out when there is none.
  • Description: one or two sentences of what should happen, including the result a person can see.
  • Steps to verify: numbered actions. The last step says what to look at and what should be true.

No priority, no tags, no source line, and no separate test-data or expected-result field. Do not use a word such as journal or handler unless the product shows that word on screen.

Code tests

Code tests are written by testing: after the feature code, or first for a bug fix. Name the behaviors it should cover. If the repo has no test runner, ask once before one is added.

Summary

Say what is still uncovered. Do not claim complete coverage when the source is ambiguous.

Replies

Talk to the user in simple English: short sentences, everyday words, and a technical term only where it is needed, explained the first time it appears.

Say how many cases, and what is still uncovered.

Source: SKILL.md on GitHub

No rule matched.

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at a0324a3. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 4 days ago.

Activeupdated 3 weeks ago

README badge

README badge for avik-creator/agent-skills/test-cases