What problem does it solve? Autonomous research pipelines need a repeatable way to check whether candidate research directions already have competing PINN/ODE literature before investing in gap analysis. This Skill standardizes multi-direction literature scanning across PubMed and OpenAlex so each cron run produces consistent, scored scan results. ## Core Features & Use Cases - Multi-Direction Scanning: Executes PubMed targeted queries plus OpenAlex broad/cited queries across 5 rotation and 5 new exploration directions per run. - Resilient API Access: Uses a Python urllib-based utility (scripts/pubmed_utils.py) with automatic retry, BOM stripping, and JSON resilience to handle NCBI eUtils instability and environments where curl is blocked by security scanners. - Scoring and State Tracking: Applies a scoring matrix (START/CANDIDATE/POSTPONE) and updates research state files with candidate status, gate results, and staleness detection between parent queues and candidate state files. - Use Case: A scheduled autonomous research agent runs this scan each cycle to confirm which vestibular/oculomotor modeling directions remain ABSOLUTE_WHITE (no PINN/ODE competitors) before queuing them for gap analysis. ## Quick Start Ask the agent to run the v32 multi-direction scan over the five rotation directions using scripts/pubmed_utils.py and report each direction's status and score.
