skills/scavio-ai/skills/scavio/Gen Agent Trust Hub

scavio

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Mentions official vendor packages including scavio and langchain-scavio for language-specific integration.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides an interface to scrape data from 50+ third-party platforms, which exposes the agent to instructions potentially embedded in untrusted web content.
  • Ingestion points: Responses from api.scavio.dev containing content from external sites.
  • Boundary markers: Instruction templates do not include specific delimiters to isolate scraped text.
  • Capability inventory: Data fetching capabilities via API endpoints.
  • Sanitization: Data is provided in structured JSON format, but textual values within the payload are not sanitized for injection patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:42 PM
Security Audit — agent-trust-hub — scavio