File naming, versioning, and documentation that prevents chaos
Set up a research project the right way from day one: consistent folder structure, explicit naming conventions, versioning rules, and minimal documentation so "future you" (or collaborators) can understand the project without a tour.
Create a skill called "Research Project Bootstrapper". Ask for: - Project type (if unknown: unspecified) - Collaboration setting (solo/team; if unknown: unspecified) - Data sensitivity (human subjects? proprietary?) to influence naming/sharing advice - Preferred naming convention components (date-first vs descriptor-first) Output: 1) Folder structure proposal. 2) Naming convention rules + examples. 3) Versioning rules (semantic vs v01 vs date-based). 4) Minimal docs: README + data dictionary + changelog templates. Rules: - If a detail is unknown, mark it as unspecified rather than guessing.
Unclear file naming and versioning creates a backlog of unorganized content.
This recipe sets up the scaffolding before the mess starts — or cleans up
an existing mess into something navigable.
De-duplicate, screen, and log decisions without losing your mind
A librarian-friendly, researcher-friendly pipeline for evidence synthesis. Import citations from multiple databases, de-duplicate, set screening rules, track decisions, and output counts plus audit logs for transparency and reproducibility.
Keep ORCID and researcher profiles current without repeated re-entry
Name ambiguity, no single place to track outputs, and the same information entered across multiple systems. This recipe builds a practical profile maintenance workflow centered on ORCID with a sync cadence so profiles don't drift.
Wikipedia-grade AI pattern removal
Comprehensive AI writing cleanup based on Wikipedia's WikiProject AI Cleanup guidelines. Catches 24+ distinct patterns including inflated symbolism, em dash overuse, rule of three, copula avoidance, and sycophantic tone.
Real sources, named experts, actual quotes
Deep research that finds primary sources with named individuals, community sentiment from Reddit/HN/X, and news coverage. No summaries of summaries — actual quotes with URLs.