How do I install the Test skill?

Run npx skillmds@latest add lantisprime/test in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Test?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

lantisprime (@lantisprime) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Test by lantisprime — Agent Skill

Name: Test
Author: lantisprime

Test

Use this skill during Phase 5 to execute functional tests, record results, log defects, and validate UX conformance. Logs defects to Git Issues when available, otherwise to markdown (or JSON, per config) under .claude/sdlc/defects/. Reports code coverage against the configured threshold. Triggered after build is signed off and before deploy. Also trigger when the user says "run tests", "report coverage", "log defects", or "validate UX".

lantisprime 1b33193 3.0 KB Updated Sep 17, 2026

File contents

Test (Phase 5)

Execute the tests defined in Design, report results, route defects.

Prerequisite

.claude/sdlc/gates/build-<task-slug>.md must exist and be signed.

Step 1 — Execute functional tests

Run the test runner configured in config/tools.json. Capture:

Pass/fail per test case (tied back to REQ IDs via the test-case metadata)
Coverage (vs. configured threshold; default 80% on modified code)
Duration, flakiness if observed

Write the execution report to .claude/sdlc/test/<task-slug>-report.md.

Step 2 — Log defects

For each failure, create a defect record:

If Git + Issues detected (per .claude/sdlc/env.json): create an issue with labels defect, severity:<level>, phase:test, and link to the test case and REQ IDs.
Otherwise: write .claude/sdlc/defects/<task-slug>/<defect-id>.md using templates/defect.md. Use JSON if artifact_format_fallback is set to json.

Every defect references at least one REQ ID and at least one test case.

Step 3 — UX conformance

If the build produced frontend output, run the UX conformance checklist:

Visual comparison against mockups (screenshots under .claude/sdlc/test/ux/<task-slug>/)
Spacing, palette, typography, component usage, state handling
Accessibility basics (contrast, keyboard nav, ARIA)

Record pass/fail items in the report. UX failures are defects, logged the same way.

Step 4 — Coverage gate

If coverage on modified code is below threshold, the phase fails. Either:

Add missing tests (but only for modified functions — see Phase 4 rule), or
Record an explicit waiver in the report with justification, requiring human gate to accept.

Human gate

Summarize: tests run, pass/fail, coverage on modified code, defects logged, UX status. Sign-off → .claude/sdlc/gates/test-<task-slug>.md.

What this skill must NOT do

Do not silently skip failing tests.
Do not inflate coverage by adding tests for unmodified code.
Do not mark a UX failure as "cosmetic" without human confirmation.

References

templates/defect.md
templates/gate.md
docs/SDLC.md Test

Next step hint

After writing the gate file, pipe the next_suggestions conditions to skills/_shared/next-hint.sh and print any output:

printf '%s\n' \
  'test_gate_signed|run /deploy to produce a deployment proposal' \
  'pending_signoff_for_current_user|write your sign-off at sign-offs/<REQ-ID>-<role>.md, then /deploy when all roles are covered' \
  | bash skills/_shared/next-hint.sh

Print any output verbatim. If the script outputs nothing, add nothing.

---name: testdescription: Use this skill during Phase 5 to execute functional tests, record results, log defects, and validate UX conformance. Logs defects to Git Issues when available, otherwise to markdown (or JSON, per config) under .claude/sdlc/defects/. Reports code coverage against the configured threshold. Triggered after build is signed off and before deploy. Also trigger when the user says "run tests", "report coverage", "log defects", or "validate UX".---# Test (Phase 5)Execute the tests defined in Design, report results, route defects.## Prerequisite`.claude/sdlc/gates/build-<task-slug>.md` must exist and be signed.## Step 1 — Execute functional testsRun the test runner configured in `config/tools.json`. Capture:- Pass/fail per test case (tied back to REQ IDs via the test-case metadata)- Coverage (vs. configured threshold; default 80% on modified code)- Duration, flakiness if observedWrite the execution report to `.claude/sdlc/test/<task-slug>-report.md`.## Step 2 — Log defectsFor each failure, create a defect record:- **If Git + Issues detected** (per `.claude/sdlc/env.json`): create an issue with labels `defect`, `severity:<level>`, `phase:test`, and link to the test case and REQ IDs.- **Otherwise:** write `.claude/sdlc/defects/<task-slug>/<defect-id>.md` using `templates/defect.md`. Use JSON if `artifact_format_fallback` is set to `json`.Every defect references at least one REQ ID and at least one test case.## Step 3 — UX conformanceIf the build produced frontend output, run the UX conformance checklist:- Visual comparison against mockups (screenshots under `.claude/sdlc/test/ux/<task-slug>/`)- Spacing, palette, typography, component usage, state handling- Accessibility basics (contrast, keyboard nav, ARIA)Record pass/fail items in the report. UX failures are defects, logged the same way.## Step 4 — Coverage gateIf coverage on modified code is below threshold, the phase fails. Either:- Add missing tests (but *only* for modified functions — see Phase 4 rule), or- Record an explicit waiver in the report with justification, requiring human gate to accept.## Human gateSummarize: tests run, pass/fail, coverage on modified code, defects logged, UX status. Sign-off → `.claude/sdlc/gates/test-<task-slug>.md`.## What this skill must NOT do- Do not silently skip failing tests.- Do not inflate coverage by adding tests for unmodified code.- Do not mark a UX failure as "cosmetic" without human confirmation.## References- `templates/defect.md`- `templates/gate.md`- `docs/SDLC.md` Test## Next step hintAfter writing the gate file, pipe the `next_suggestions` conditions to `skills/_shared/next-hint.sh` and print any output:```bashprintf '%s\n' \ 'test_gate_signed|run /deploy to produce a deployment proposal' \ 'pending_signoff_for_current_user|write your sign-off at sign-offs/<REQ-ID>-<role>.md, then /deploy when all roles are covered' \ | bash skills/_shared/next-hint.sh```Print any output verbatim. If the script outputs nothing, add nothing.

Frequently asked questions