The Agent Skills Directory

[SAFE]: No malicious patterns or vulnerabilities were detected in the skill code or instructions.
[COMMAND_EXECUTION]: The skill uses the Bash tool to interact with the local file system and project structure. This is the intended behavior for managing .claude/agents/ files.
[PROMPT_INJECTION]: The skill processes user-provided arguments to define agent behaviors but includes instructions for critical thinking and verification (skepticism, confidence thresholds), which serves as a defensive measure against malicious inputs.
[DATA_EXPOSURE]: The skill reads and writes files within the specific .claude/agents/ directory. No evidence of unauthorized file access or credential harvesting was found.

custom-agent