gerador-de-agentes

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill takes user-provided task descriptions, tool names, and skill sets to generate structured agent definitions.
  • Ingestion points: The 'context' and 'tarefa' sections in SKILL.md define how user input (tasks, tools, skills) is used to create agent specs.
  • Boundary markers: There are no instructions for the agent to use delimiters or to treat user-provided tasks as data rather than instructions, which could lead to malicious tasks influencing the generated agent's persona.
  • Capability inventory: The skill generates agents intended to use tools and skills. While the generator skill itself has no sensitive tool access, the generated agents are designed for action, creating a potential path for injected instructions to influence tool execution.
  • Sanitization: The skill lacks any requirement to sanitize or validate the user-provided strings before inserting them into the 'goal', 'persona', or 'description' fields of the generated agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 06:24 PM
Security Audit — agent-trust-hub — gerador-de-agentes