What problem does it solve?
It generates a sharp, visual, technical explainer for a specific recent mechanism, so you can understand how something works without reading scattered sources or guessing the core move.
Core Features & Use Cases
- Deterministic topic selection: Picks the newest relevant mechanism from recent articles, paper picks, or memory logs when you don’t provide an explicit topic.
- Triangulated research: Runs three targeted WebSearch queries to extract the core mechanism, failure modes, and concrete quantitative anchors with primary-source emphasis.
- Replica hero image with safe fallback: Attempts to generate a labeled schematic via Replicate when
REPLICATE_API_TOKEN is set; otherwise it reliably ships a text-only explainer and logs the skip reason.
- Structured, reusable output: Saves a 600–1000 word explainer (with sources) to
articles/ and notifies with the key idea line and read link.
Quick Start
Run the skill with an explicit topic like Technical Explainer var="entropy trajectory reasoning" to produce articles/explainer-<today>.md including a key idea, step-by-step mechanism, numbers with sources, and a notification.
