inboundfy-pesquisa-acervo
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is instructed to fetch and analyze external web content and local project files. 1. Ingestion points: Mandatory web research (Step 2 of Flow in SKILL.md) and user-supplied files such as bruto.md and processado.md. 2. Boundary markers: Absent; the instructions do not specify the use of delimiters or 'ignore' directives to separate external content from agent instructions. 3. Capability inventory: The agent uses web search tools and has permissions to read local configuration files and write to pesquisa.md. 4. Sanitization: Absent; no requirements are provided for sanitizing, escaping, or validating the integrity of external information before it is processed.
Audit Metadata