smoke-testing

Enforce post-deploy smoke tests verifying operator-visible outcomes with ground-truth diagnostics.

6|Updated May 29, 2026
One-click install
npx skills add https://github.com/fusebase-dev/fusebase-flow --skill smoke-testing-fusebase-dev
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: smoke-testing
Source: https://github.com/fusebase-dev/fusebase-flow/tree/main/flow-skills/smoke-testing
Command: npx skills add https://github.com/fusebase-dev/fusebase-flow --skill smoke-testing-fusebase-dev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the risk of teams mistaking pre-deploy static checks (like lint, unit tests, or exit code validation) for successful smoke testing, which often lets production bugs slip through because true smoke testing requires proving operator-visible outcomes work on the deployed surface with supporting ground-truth diagnostics.

Core Features & Use Cases

  • Strict Smoke Contract Definition: Guides Product Owners to write smoke tests that require verifiable operator-visible outcomes and named ground-truth diagnostic surfaces, rather than relying on weak pre-deploy signals.
  • End-to-End Smoke Execution: Provides AI Developers with a step-by-step process for running post-deploy smoke checks, including UI interaction testing, backend diagnostic cross-checking, adversarial falsification probes, and structured evidence collection.
  • Use Case: When deploying a new client-facing help-desk feature, this Skill ensures the feature actually works for end users in production, rather than only passing pre-deploy automated checks that miss runtime or integration bugs.

Quick Start

Use the smoke-testing skill to define a post-deploy smoke test contract for your new feature and execute the required smoke checks after deployment to confirm operator-visible outcomes work correctly with supporting ground-truth diagnostics.

Frequently Asked Questions about smoke-testing

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

FAQPage Schema
Why does passing pre-deploy static checks still let production bugs slip through after deployment?

Static pre-deploy checks like lint and unit tests do not verify operator-visible outcomes on the live surface, allowing integration or runtime bugs to slip through. True smoke testing requires post-deploy verification with ground-truth diagnostics to confirm features work in production.

How do I write a smoke test contract that verifies operator-visible outcomes?

Define a smoke test contract by specifying verifiable operator-visible outcomes and naming ground-truth diagnostic surfaces. Include adversarial falsification probes and stable UI locators to ensure the deployed feature is actually exercised rather than relying on weak pre-deploy signals.

How do I run post-deploy smoke checks for application deployments?

Run post-deploy smoke checks by executing UI interaction tests, cross-checking backend diagnostics, performing adversarial falsification probes, and persisting structured evidence. This step-by-step process confirms operator-visible outcomes work correctly with supporting ground-truth diagnostics.

What's the difference between static validation and true smoke testing?

Static validation confirms code passes pre-deploy automated checks like lint and unit tests, whereas true smoke testing proves operator-visible outcomes function on the deployed surface. Static validation misses runtime bugs that post-deploy ground-truth diagnostics are designed to catch.

Can I use smoke testing for both user-facing and operator-facing feature changes?

Yes, smoke testing supports both user-facing and operator-facing feature changes. Product Owners can draft smoke contracts for either surface type, ensuring verifiable outcomes and ground-truth diagnostics are defined regardless of the audience interacting with the deployed feature.

How do I persist structured evidence from post-deploy smoke tests for auditability?

Persist structured evidence by collecting outputs from UI interactions, backend diagnostic cross-checks, and adversarial falsification probes during post-deploy execution. This structured evidence provides auditability by proving operator-visible outcomes were verified against ground-truth diagnostic surfaces.