viasrijan

Use when agent sessions generate long histories, context window limits are hit, or you need to preserve decisions and artifact trails…

Use when output quality drops or you switch tasks — feed agents the right information at the right time.

Use when diagnosing or optimizing LCP, INP, CLS — Core Web Vitals at the 75th percentile (LCP ≤2.5s, INP ≤200ms, CLS ≤0.1).

Use when making CTO decisions — tech debt, team scaling, architecture, and DORA metrics.

Use when running autonomous research — plan, search, read, and synthesize into reports.

Use when making architectural decisions, changing APIs, or shipping features — write Architecture Decision Records and API docs that…

Use when stakes are high — adversarial fresh-context review of every non-trivial decision.

Use when building or auditing frontend UI — components, design systems, or avoiding generic AI aesthetics.

Use when building UI components — design systems, state, responsive, a11y.

Use when making any code change — trunk-based development, atomic commits, change sizing (~100 lines), and the commit-as-save-point pattern.

Use when you have a rough concept that needs divergent and convergent exploration.

Use when implementing changes touching multiple files — thin vertical slices, feature flags, rollback-friendly.