The Agent Skills Directory

[SAFE]: The skill is purely instructional and documentation-based, providing templates and architectural guidance for business automation.
[EXTERNAL_DOWNLOADS]: References official and well-known services such as the n8n Docker image and the Anthropic Claude API. These are documented for the user's manual setup and do not involve silent or untrusted downloads.
[COMMAND_EXECUTION]: Provides example configurations for docker-compose and n8n node settings. These examples use industry-standard practices, such as environment variables for sensitive configuration, and are intended for the user's own deployment.
[PROMPT_INJECTION]: No patterns of prompt injection, role-play bypasses, or instruction overrides were detected. The skill's instructions are limited to business logic and automation design.
[DATA_EXFILTRATION]: No attempts to access sensitive local files or exfiltrate data were found. The skill explicitly promotes the use of human checkpoints and review gates for high-stakes decisions and financial operations.
[REMOTE_CODE_EXECUTION]: Does not contain any scripts or commands that execute remote code on the host system. It provides architectural patterns for users to build their own integrations.

agents