The Agent Skills Directory

[COMMAND_EXECUTION]: The skill leverages the x openspec command-line interface to execute project-specific tasks such as init, change, and validate. These commands are used to maintain project structure and align with the skill's stated purpose of coordinating coding collaboration.
[DATA_EXPOSURE]: The framework operates on local markdown files (.md) to track specifications and proposed changes. It does not attempt to access sensitive system files (e.g., .ssh, .aws) or credentials.
[SAFE]: Analysis across all 10 threat categories confirms the absence of prompt injection, obfuscation, persistence mechanisms, or unauthorized remote code execution. The workflow relies on user-initiated commands and standard project management practices.

openspec