The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were identified. The workflow is designed for productivity and uses standard platform tools for codebase analysis and planning.- [COMMAND_EXECUTION]: The skill utilizes the internal Agent tool to spawn specialized sub-agents (analyzer, architect) for modular tasks. These operations are scoped to the project environment and do not involve arbitrary shell commands or remote execution.- [DATA_EXFILTRATION]: While the skill analyzes the codebase, all findings are stored locally in the .spec/ directory. The skill mandates that this directory be added to .gitignore, protecting project metadata from accidental exposure to public repositories.

spec-coding