The Agent Skills Directory

[SAFE]: The skill performs legitimate documentation tasks, such as creating ADR files and reading project structure. No suspicious behaviors, obfuscation, or unauthorized data access patterns were identified.
[COMMAND_EXECUTION]: The skill uses local utility commands like find and grep to search the project directory for relevant configuration and source files (e.g., SQL schemas, models, and Docker configurations). These commands are used solely to provide context for the agent to write accurate architectural documentation and do not involve network operations or exfiltration.

docs-adr