The Agent Skills Directory

[SAFE]: The skill serves as an instructional framework for multi-agent role-play. It provides templates for various expert personas to provide feedback on user-supplied topics. The logic is purely based on synthesis and perspective-taking without performing unauthorized system operations.
[INDIRECT_PROMPT_INJECTION]: The skill features a data ingestion surface where user-provided questions are interpolated into prompts for various personas. While this creates a potential surface for instructions embedded in user data to influence the persona responses, the skill is designed for advisory purposes and includes specific guidelines for the orchestrator to maintain persona authenticity.
Ingestion points: User input is received via the RECEBER: Questao do usuario step described in section 3.1 of SKILL.md.
Boundary markers: The skill does not explicitly use advanced delimiters or "ignore previous instructions" safety wrappers for the user-provided question within its synthesis templates.
Capability inventory: The skill utilizes tools like claude-code and gemini-cli as specified in the frontmatter, but does not contain direct eval(), exec(), or arbitrary shell command execution logic in the provided instructions.
Sanitization: No explicit sanitization or filtering of user input is implemented within the orchestration logic.

multi-advisor