xp-practices

This skill provides a structured framework for adopting Extreme Programming (XP) methodologies within modern engineering teams, especially those leveraging AI-assisted workflows. It focuses on the systemic integration of core agile practices such as TDD (Test-Driven Development), Pair Programming, Ensemble Programming, and Continuous Integration. By treating these practices as a cohesive system rather than individual tasks, engineering teams can achieve a sustainable pace of delivery while maintaining high technical standards and collective code ownership. It is designed for software engineers, quality assurance professionals, and technical leads who aim to formalize their development lifecycle through disciplined, test-centric habits.

• Orchestrates agentic workflows for TDD cycles including RED-GREEN-REFACTOR phases.

• Provides standardized patterns for Driver-Navigator and Ping-Pong pair programming sessions.

• Integrates continuous integration logic with automated risk analysis and regression testing.

• Supports collective ownership via code review coordination and knowledge-sharing protocols.

• Offers practical guidance on when to apply pairing (e.g., complex code, onboarding, critical bugs) vs. when to perform independent research spikes.

• Facilitates memory management for pairing sessions, team velocity metrics, and retrospective logs.

• Start with high-value practices like TDD and CI before introducing collective ownership or small releases.

• Rotate pairing roles every 20-30 minutes to maintain energy and engagement.

• Use the provided fleet coordination schema to integrate specialized agents like qe-test-generator, qe-test-executor, and qe-code-reviewer.

• Align AI agents to handle repetitive grunt work such as regression suites, data generation, and coverage analysis, while humans focus on architectural judgment and strategy.

• Adhere to the principle that XP works as a system; avoid cherry-picking practices that might undermine the safety net of automated testing or continuous feedback loops.

• The skill maintains internal state across sessions, allowing the agent to remember patterns learned from your specific codebase and team dynamic.