assert

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADS
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill documentation provides a command to download content from an external domain that is not on the trusted sources list.
  • Evidence: curl -H "Accept: text/plain" https://fullstackrecipes.com/api/recipes/assert in SKILL.md.
  • Risk: Fetching scripts or instructions from unverified third-party domains can lead to the ingestion of malicious code or prompt injections. If the agent is instructed to follow the 'recipe' retrieved, it could perform actions dictated by the external server.
  • INDIRECT PROMPT INJECTION (LOW): The skill is designed to ingest untrusted data from an external URL.
  • Ingestion points: SKILL.md (via the curl command).
  • Boundary markers: Absent; there are no instructions to the agent to treat the fetched content as untrusted or to ignore embedded commands.
  • Capability inventory: The skill description suggests the content is for 'TypeScript assertion', implying the agent will handle or execute code based on the download.
  • Sanitization: None; the content is fetched and used directly without validation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 05:27 PM