What problem does it solve?
Turning a broad research direction into concrete, publishable ideas is slow and error-prone: researchers must survey the literature, spot gaps, check novelty, and guess which ideas are worth GPU time. This Skill automates that pipeline, producing a ranked idea report backed by landscape analysis, external LLM critique, and small pilot experiments.
Core Features & Use Cases
- Landscape Survey: Scans local paper libraries and recent literature (top venues, arXiv) to map sub-directions, gaps, and open problems before ideation.
- LLM-Augmented Brainstorming & Review: Uses an external model via Codex MCP to generate 8-12 candidate ideas, then applies devil's-advocate critique and novelty checks to filter them down.
- Parallel Pilot Experiments: Runs minimal GPU experiments (with strict time and GPU-hour budgets) for the top 2-3 ideas and re-ranks them based on empirical signal.
- Use Case: A researcher says "find ideas on sample efficiency of offline RL with image observations" and receives an IDEA_REPORT.md with ranked hypotheses, novelty scores, pilot results, eliminated dead ends, and a suggested execution order.
Quick Start
Ask the assistant to run the aris-idea-creator skill with a specific research direction such as "factorized gap in discrete diffusion language models" to generate a ranked idea report.
