The Agent Skills Directory

[SAFE]: The skill serves as a documentation template and instructional guide for organizing project-specific automation. It does not perform any network operations, access sensitive files, or execute unauthorized commands.
[PROMPT_INJECTION]: The skill acts as a factory for creating new skills based on session history. While this introduces an ingestion surface for potentially untrusted context, the skill mitigates this risk by providing a mandatory quality checklist that explicitly instructs the agent to exclude secrets, hardcoded paths, and environment-specific values.
[COMMAND_EXECUTION]: The skill recommends a directory structure that includes a placeholder for automation scripts. This guidance is organizational in nature and does not involve the skill executing arbitrary or hidden commands.

skillforge