# Testing Regression Strategy > Risk-based regression suite curation for release gating under time and budget limits. Use when teams must decide what runs always versus conditionally by impact and risk; do not use for implementing a single test type in isolation. Use when this capability is needed. - Skill: `tomevault-io/testing-regression-strategy` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/testing-regression-strategy` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/testing-regression-strategy/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/testing-regression-strategy --- # Testing Regression Strategy ## Overview Use this skill to select regression coverage that maximizes risk reduction within execution constraints. ## Scope Boundaries - Use when not all tests can run every change and selection policy is required. - Typical requests: - `Optimize regression coverage under CI time budget.` - `Define mandatory versus conditional suites by risk.` - `Map change impact to regression gates.` - Do not use when: - The task is implementing one specific test case only. - The task is observability/alert policy design (`observability-*`). ## Inputs - Release risk profile and CI budget - Test inventory with cost, flakiness, and detection value - Change-impact and component criticality model ## Outputs - Tiered regression policy (always-run, conditional, periodic) - Decision record for suite selection logic and trade-offs - Verification checklist and gap register ## Workflow 1. Classify system areas by business and technical risk. 2. Quantify each suite by execution cost and detection value in `assets/regression-tiering-template.csv`. 3. Compare selection policies and choose one with rationale. 4. Define trigger rules for conditional suites. 5. Publish policy, residual risk, and review cadence. ## Quality Gates - High-risk areas are always covered by mandatory gates. - Selection policy is explainable and auditable. - Budget trade-offs are explicit, not implicit. - Policy includes periodic recalibration triggers. ## Failure Handling - Stop when high-risk areas are uncovered by mandatory tests. - Escalate when budget and required risk coverage conflict. ## Bundled Resources - `references/trigger-and-examples.md`: trigger patterns, anti-patterns, and deliverable expectations. - `assets/regression-tiering-template.csv`: quick-start matrix for always/conditional/nightly suite selection. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/kentoshimizu) — claim your Tome and manage your conversions.