What problem does it solve? It answers feasibility questions that research alone cannot resolve by building small, disposable prototypes, so you avoid committing to an approach that turns out to be unworkable. ## Core Features & Use Cases - Decompose and Prioritize: Breaks an idea into 2-5 independent feasibility questions framed as Given/When/Then, ordered by risk so the idea-killing question runs first. - Research and Build Loop: Surfaces competing libraries or approaches, picks one, then builds a runnable CLI, HTML page, or small server per spike in a spikes/NNN-name/ directory. - Verdicts and Comparisons: Closes each spike with a VALIDATED, PARTIAL, or INVALIDATED verdict, and produces head-to-head comparisons for parallel approaches (e.g., pdfjs vs camelot). - Use Case: Before building a streaming chat feature, spike whether websocket token streaming reaches the client in under 100ms, then throw the prototype away once the verdict is in. ## Quick Start Spike whether websocket streaming can deliver LLM tokens to the browser in under 100ms before I commit to this architecture.
