The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.check_output in scripts/scan_environment.py to identify available tools (git, python3, node, etc.) and project status. These are standard environment discovery patterns for developer tools and do not involve untrusted input or risky execution chains.- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill maintains project state in a local .kkirikkiri/ directory. There are no network operations or external URL requests detected in the scripts or instructions. All data handling is local to the user's workspace.- [PROMPT_INJECTION]: The instructions focus on structured delegation and task management. No patterns were found attempting to bypass AI safety filters, override system instructions, or extract system prompts.- [DYNAMIC_EXECUTION]: The init_shared_memory.py script generates markdown files for team coordination. These are static text files and are not executed as code, posing no risk of dynamic code injection.

kkirikkiri