context-driven-testing

Context-driven testing is a methodological framework for software agents designed to replace rigid, one-size-fits-all testing dogma with flexible, risk-aware decision-making. This skill empowers agents to evaluate the unique landscape of a project—including team expertise, project stage (startup vs. enterprise), safety-critical requirements, and resource constraints—to tailor the testing approach dynamically. By prioritizing investigation over mere test script execution, the agent ensures that testing efforts are concentrated where they provide the most value, reducing wasteful overhead and increasing the speed of finding critical bugs.

• Adaptive decision support based on project context, goals, and technical risks.
• Utilizes Rapid Software Testing (RST) heuristics such as SFDIPOT (Structure, Function, Data, Interfaces, Platform, Operations, Time) to categorize testing focus.
• Facilitates agent-assisted exploratory testing, recommending automated tools like performance scanners or security auditors only when they align with current project needs.
• Supports intelligent fleet coordination, allowing the agent to spawn specialized sub-agents based on the analyzed environment.
• Enables documentation of discoveries and learning patterns rather than static, pre-written test plans, fostering an evolving knowledge base.
• Integrates risk assessment directly into the testing loop, prioritizing high-risk code paths for intensive coverage.

Use this skill when initiating a new project, questioning existing testing dogmas, or pivoting your strategy to handle shifting project constraints. The agent acts as a lead quality architect, accepting inputs like project stage, compliance requirements, and business goals to output a customized testing strategy. Constraints are handled by the agent's logic, which prevents over-engineering in small, fast-paced startups while ensuring rigor in regulated enterprise environments. It is ideal for developers and QE engineers who need to pair human-like judgment with AI-driven execution to maintain high velocity and software stability simultaneously.