The Agent Skills Directory

[SAFE]: The skill defines organizational standards for AI agent definitions and includes a local Python script for validating README indexes. No malicious patterns, network operations, or unauthorized data access were found.
[COMMAND_EXECUTION]: The skill references a CLI tool 'wagents' for validation. This appears to be a local development utility intended for use within the agent's environment to verify configuration compliance.
[DATA_EXFILTRATION]: File access is restricted to local agent configuration directories (e.g., '.claude/agents/') for the purpose of maintaining internal documentation and metadata. No network calls or exfiltration vectors are present.

agent-conventions