Jait

Prove bug fixes with fail-then-pass validation. Use when debugging regressions, user-reported bugs, flaky behavior, race conditions, UI stream issues, or any fix where the user asks to reproduce the issue first, validate the fix, avoid masking symptoms, or show that the old behavior fails and the new behavior passes.

System design, task decomposition, and technical planning. Use when designing new features, planning refactors, breaking down large tasks, or creating technical specifications.

Systematic code review covering correctness, security, performance, and maintainability. Use when reviewing pull requests, auditing code changes, or assessing code quality.

Structured data exploration, transformation, and analysis. Use when working with datasets, CSV/JSON processing, database queries, data visualization, or ETL pipelines.

Systematic root cause analysis and bug fixing. Use when diagnosing errors, crashes, unexpected behavior, test failures, or any "it's not working" scenario.

Systematic web research with source triangulation, structured synthesis, and citations. Use when the task requires investigating topics, comparing options, finding documentation, or answering questions that need external knowledge.

Performance analysis, profiling, and optimization. Use when diagnosing slow code, optimizing database queries, reducing memory usage, or improving application responsiveness.

Security-focused code analysis covering OWASP Top 10, authentication, authorization, and data protection. Use when auditing code for vulnerabilities, reviewing security-sensitive changes, or hardening an application.

Generate comprehensive tests with good coverage strategy. Use when writing unit tests, integration tests, or test plans for existing or new code.