Ready-to-use prompts for coding agents. Copy a complete task as-is; the agent reads available project context, uses safe defaults, and keeps implementation boundaries and acceptance checks explicit.
Use the prompt as-is, add real project context when available, and verify the running product, tests, and release evidence yourself.
Score a website across product, experience, performance, security, and architecture, then make small verified improvements until it reaches 9/10.
Build the information hierarchy, conversion path, responsive implementation, and release checks around one business goal.
Analyze visual and interaction evidence first, then recreate it responsively with the project's existing stack.
Extract tokens, states, and reuse boundaries from a design specification without over-abstracting the page.
Use a traceable baseline to locate the regression and repair it without changing the intended design behavior.
