skills/dydoc/skills/grilling/Gen Agent Trust Hub

grilling

Pass

Audited by Gen Agent Trust Hub on Sep 12, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to autonomously 'dispatch a sub-agent' to find facts from the environment (filesystem, tools) based on user-provided plans or decisions. This creates a vulnerability surface where instructions embedded in the user input or in the external files the agent reads could manipulate agent behavior.
  • Ingestion points: User-supplied plan, decision, or idea (SKILL.md).
  • Boundary markers: Absent; the instructions lack delimiters or warnings to ignore embedded commands in environmental data.
  • Capability inventory: Sub-agent dispatch, filesystem access, and generic tool execution (SKILL.md).
  • Sanitization: Absent; no validation, escaping, or filtering of environmental facts or user input is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 12, 2026, 06:28 AM
Security Audit — agent-trust-hub — grilling