What problem does it solve?
Daily research news feeds produce large volumes of raw items across arXiv, GitHub, HuggingFace, and social sources, but turning them into actionable research ideas requires manual clustering, gap analysis, and citation tracking that is slow and error-prone.
Core Features & Use Cases
- Deterministic aggregation: A Python helper reads every news-results-*.json file, normalizes scores across sources, dedupes by canonical id (arXiv ID, repo URL, title hash), and groups items into preliminary clusters.
- Grounded ideation: The workflow refines clusters, proposes 2-3 idea seeds per cluster with rationale, first experiment, risk, and confidence, and enforces anti-hallucination rules requiring every claim to cite a candidate item.
- Structured outputs: Produces a human-readable idea_briefing.md narrative and a machine-readable seeds.json under Ideation/proactive/<date>/, with append-only history.
- Use Case: After your daily news feed runs, invoke this Skill to get a dated briefing of 5-10 thematic clusters with concrete, citation-backed research ideas, optionally cross-checked for novelty.
Quick Start
Turn today's news feed results into a clustered idea briefing with grounded research seeds saved under my project's Ideation folder.
