The Agent Skills Directory

[SAFE]: The skill implements a structured Q&A workflow to generate agent configuration files in markdown and YAML formats. It acts as a design assistant rather than executing complex logic itself.- [SAFE]: Security best practices are enforced within the instructions, specifically requiring the agent to configure permission.skill with a default-deny policy ("*": "deny") and explicit whitelisting for all generated agents.- [SAFE]: The skill includes a research phase that encourages the use of web search tools to find current best practices, which is an expected and benign function for a development architect tool.- [SAFE]: No evidence of prompt injection, credential harvesting, or unauthorized network communication was found in the instructions or referenced files.

agent-architect