periplus-dataset-research

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from search tools and web discovery, creating a surface for indirect prompt injection.\n
  • Ingestion points: The skill ingests data from external URLs, acquired pages, and Periplus corpus search results (SKILL.md).\n
  • Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when processing external sources.\n
  • Capability inventory: The skill executes SQL queries and performs link following (SKILL.md).\n
  • Sanitization: Instructions include validation steps for duplicates and conflicting identities to ensure data integrity.\n- [DYNAMIC_EXECUTION]: The skill involves generating and saving SQL queries for execution against the Periplus dataset (SKILL.md).\n- [EXTERNAL_DOWNLOADS]: The skill references documentation and operational guides from the external domain docs.periplus.dev.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 04:34 AM