Skip to content
Agentmart
Submit a tool

coding-unit-test

Guides unit-test work in any language.

Skill by andresgongora

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/andresgongora-coding-unit-test

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Guides unit-test work in any language. Load for unit tests, TDD, testability, coverage, or test-framework choice. Skip for tiny glue and unstable spikes that need no test-first ritual, or for test infrastructure and CI configuration.

Use cases: testing, devops

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/coding-unit-test

What's inside

Skill name
coding-unit-test
Description
Guides unit-test work in any language. Load for unit tests, TDD, testability, coverage, or test-framework choice. Skip for tiny glue and unstable spikes that need no test-first ritual, or for test infrastructure and CI configuration.
License
MIT
Instructions
74 lines in SKILL.md · view on GitHub

Files (2)

  • README.md
  • SKILL.md

SKILL.md (first lines)

Discipline for useful, deterministic tests. No unnecessary tooling. No internal coupling.

## Outcome

Prove stable behavior. Follow repo idiom. Never bind to implementation internals.

## Inspect First

Before writing tests:

- test command and framework in use.
- test locations and file naming convention.
- fixtures, fakes, mocks, helpers in use.
- assertion idiom and async/error handling.
- nearby tests for the same boundary.

Existing setup: do NOT add framework, runner, layout, or assertion library unless it genuinely cannot test the contract.

No setup exists:

- material, repo-wide framework choice: ask.
- clear ecosystem default: choose it, state choice.
- no clear default: choose one sane default, state choice; use root `tests/` unless language or tooling convention gives stronger evidence.

## Design

1. Name stable module boundary.
2. State contract: inputs, outputs, effects, invariants, failures.
3. Choose honest test level:
   - pure unit.
   - unit with fake at adapter seam.
   - integration test.
   - boundary too muddy: improve design before forcing test.
4. Cover behavior branches, errors, and regressions with focused named cases.
5. Assert observable results only. Private calls, internal structure, incidental order: untested.

Prefer: pure functions, explicit dependencies, adapter-isolated effects, deterministic time/randomness, narrow fakes.

Painful setup signals bad boundary. Improve boundary before adding mock scaffolding.

Reject:

- booting half system for local logic.
- mocks that restate implementation or lack contract.
- timing/network/filesystem/randomness/shared-state flake sources.
- snapshots/goldens where explicit assertions say more.

## TDD and Coverage

- Stable, valuable seam: test before, with, or after change — favor early feedback.
- Tiny glue, exploratory code, unstable interface: no ritual test-first.
- Coverage measures gaps; never chase percentage with low-value tests.
- Intentional contract change: update affected tests — do not preserve obsolete expectations.

## Validate

Run narrow relevant tests first, then repo-standard test command when proportionate. Report exact command, result, and unrun scope. Report coverage metric only from coverage command; inspection may identify untested paths.

## Escalate

Ask before:

- competing framework conventions within the same repo.
- large testability refactor.
- unclear or conflicting test policy.

## Boundaries

- Stable behavior deserves tests; every glue line does not.
- Integration behavior stays labelled integration.
- Repo forbids or limits tests: obey.
- Not for test infrastructure or CI configuration.

From andresgongora/agent-config, MIT.

More from andresgongora/agent-config

Signals

Stars
0
Forks
1
Open issues
0
Last commit (this folder)
Aug 30, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from andresgongora

Cite this page

coding-unit-test skill. Agent Mart. https://agentmart.dev/skills/andresgongora-coding-unit-test (last updated 2026-10-24)