OverviewHistoryStatsSecurity
Security
Independent checks from skills.sh's audit partners.
Independent checks from skills.sh's audit partners.
Independent checks from skills.sh's audit partners.
Warn5 providerslatest audit May 29, 2026
Independent checks from skills.sh's audit partners.
The 'deep-research' skill is a sophisticated and highly disciplined 13-agent pipeline designed for rigorous academic research. It incorporates exceptional security and integrity controls, such as three-index API triangulation (Semantic Scholar, Crossref, OpenAlex) to detect hallucinated citations, temporal integrity logic to verify event timelines, and 'pattern protection' rules to prevent model drift. The skill uses standard utilities and well-known academic APIs for legitimate verification purposes, demonstrating a high degree of defensive design against common AI agent failure modes.
Detected behaviors
No alerts
2 issues
7/41 files flagged
Score: 93/100 · 2 sections analyzed