skills/kinhluan/skills/sota-survey/Gen Agent Trust Hub

sota-survey

Pass

Audited by Gen Agent Trust Hub on May 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches paper metadata from the Semantic Scholar API and HuggingFace Hub. These are well-known services for academic and machine learning research data.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external research papers.\n
  • Ingestion points: Research paper metadata and full-text content are ingested during the triage workflow.\n
  • Boundary markers: The skill does not define specific delimiters for separating paper content from internal instructions.\n
  • Capability inventory: Actions include writing to the local file system (/research/notes/sota-matrix.md) and executing network requests via curl.\n
  • Sanitization: No sanitization or validation of the ingested paper content is performed.\n- [SAFE]: No patterns of obfuscation, persistence, or privilege escalation were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
May 21, 2026, 08:12 AM
Security Audit — agent-trust-hub — sota-survey