executing-integration-tests

Execute integration test suites in a sandbox and emit runner-backed PASS, FAIL, or ERROR verdicts.

Updated Apr 26, 2026
One-click install
npx skills add https://github.com/Jin9/skillify-foundation --skill executing-integration-tests-jin9
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: executing-integration-tests
Source: https://github.com/Jin9/skillify-foundation/tree/main/treasury/executing-integration-tests
Command: npx skills add https://github.com/Jin9/skillify-foundation --skill executing-integration-tests-jin9

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? After implementation, teams need a trustworthy integration-level (SIT) gate that proves cross-service flows work in a wired-up environment, without relying on LLM-asserted passes or leaking state between runs. ## Core Features & Use Cases - Runner-backed execution: Runs the integration suite through the CI/test runner in a sandbox, capturing per-flow results with reproduction evidence. - Teardown verification: Confirms the sandbox tears down cleanly; a dirty teardown escalates as an exception rather than a pass. - Strict gate policy: Emits PASS only when zero flows failed and teardown is clean; uncertain results count as FAIL, and defects route to a fixer with severity classification. - Use Case: After a banking feature is implemented, run the SIT suite against the sandbox to produce an auditable pass/fail gate with defect evidence for a human verification layer. ## Quick Start Run the integration test suite against my sandbox environment and give me the pass/fail verdict with any defects found.

Frequently Asked Questions about executing-integration-tests

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run integration tests after implementation is complete?▼

Provide a test roster or built artifacts plus a reachable sandbox target environment and an idempotency key. The suite executes through the CI/test runner, and you receive a PASS, FAIL, or ERROR verdict with per-flow results and defect evidence.

What is the difference between integration tests and unit or e2e tests?▼

Integration tests (SIT) verify cross-service flows against a wired-up system in a sandbox. Unit tests cover isolated backend or frontend components, while e2e tests cover full user journeys; this skill handles only the integration level.

Can this skill fix failing integration tests automatically?▼

No. It executes and reports only; it never writes or fixes tests or production code. Failures are routed to a fixer such as progressive-bug-hunter, and the verdict feeds a human verification layer.

Why did my integration run fail even though all tests passed?▼

A dirty teardown causes a FAIL with teardown_ok set to false. Leaked fixtures, open connections, or residual state escalate as an exception because they corrupt later runs and hide defects.

What happens when integration test results are flaky or uncertain?▼

Uncertain or masked results are treated as FAIL under the banking-grade gate policy. Non-deterministic flows are surfaced in the defects list and flagged for human verification rather than passed by luck.