scansci-sort

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external academic paper identifiers and metadata, which provides an attack surface for instructions embedded in the data.
  • Ingestion points: Large lists of academic paper identifiers and metadata (SKILL.md).
  • Boundary markers: No delimiters or "ignore instructions" warnings are specified to separate untrusted data from agent instructions.
  • Capability inventory: The workflow references probing sources and bucketing files, which suggests the agent has network and file system access (SKILL.md).
  • Sanitization: The instructions do not define any validation or sanitization requirements for the ingested paper data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 01:12 AM
Security Audit — agent-trust-hub — scansci-sort