strategy-operations

Plans test placement, environments, data, flakiness handling, and CI gating for test strategy.

Updated Jun 24, 2026
One-click install
npx skills add https://github.com/Hakkadaikon/hymme --skill strategy-operations-hakkadaikon
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: strategy-operations
Source: https://github.com/Hakkadaikon/hymme/tree/main/skills/strategy-operations
Command: npx skills add https://github.com/Hakkadaikon/hymme --skill strategy-operations-hakkadaikon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Deciding where tests run, what data they consume, and how to keep CI green over time is often left to ad-hoc choices, leading to flaky suites, polluted shared environments, and gates that nobody trusts. This Skill provides the operational half of a test strategy: shift-left/shift-right placement, environment and data strategy, flakiness quarantine loops, and CI gating with time budgets. ## Core Features & Use Cases - Shift-left / Shift-right Placement: Assign each verification to the cheapest stage that can catch it, from design reviews and TDD on the left to production monitoring, synthetic monitoring, and canary releases on the right. - Environment & Data Strategy: Allocate tests across local/dev, Testcontainers, staging, and ephemeral environments, and manage fixtures, builders, seeding, generated data, and masked production data for independence and reproducibility. - Flakiness & CI Governance: Run a detect → quarantine → fix → restore loop for flaky tests, and set risk-based coverage/mutation thresholds, parallelization, and execution time budgets as merge gates. - Use Case: A team's CI takes 90 minutes and flaky tests erode trust. Use this Skill to move logic coverage down to fast layers, quarantine flaky tests with owners and deadlines, and enforce risk-based gates with a time budget. ## Quick Start Ask the AI to design the operational test strategy for your project, covering where each test runs, what data it uses, how flaky tests are quarantined, and what CI gates and time budgets apply.

Frequently Asked Questions about strategy-operations

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

FAQPage Schema
How do I decide which tests run in which environment?

Sort tests by confidence requirement: logic coverage that test doubles can handle goes to fast local/dev layers, while configuration-dependent failures go to staging or production-like environments. Use Testcontainers for real databases and queues, and ephemeral environments for tests needing isolation.

What is the difference between shift-left and shift-right testing?

Shift-left moves verification earlier into design review, linting, and TDD to fix defects cheaply before production. Shift-right uses production monitoring, synthetic monitoring, and canary releases to catch issues that only appear with real traffic and data.

How should flaky tests be handled in CI?

Detect flaky tests via pass/fail history, quarantine them from the main suite with a deadline and owner, fix the root cause such as timing or ordering dependencies, then restore them after verified stability. Never silence them with retries or permanent skips.

Should coverage thresholds be the same for all code?

No. Apply risk-based thresholds: strict coverage and mutation score minimums for critical areas like money, security, and state transitions, and looser limits for low-risk code. Uniform high thresholds force expensive tests where they add little value.

When should I use Testcontainers instead of mocks?

Use Testcontainers when a fake would diverge from real behavior, such as SQL dialects or message queue semantics, for dependencies you control. It runs real containers from unit and integration tests without needing a dedicated environment.

Why does my CI gate stop being effective over time?

Gates fail when there is no execution time budget, so slow end-to-end tests pile up and developers stop waiting for green. Set time budgets per level, push slow coverage down to faster layers, parallelize, or reduce run frequency.