The Agent Skills Directory

[SAFE]: The skill instructions focus on summarizing conversation context into 'checkpoints' to manage token limits. It explicitly warns against manipulating internal client configuration or deleting chat databases.
[SAFE]: The helper script suggest-compact.sh implements stateless logic to print a reminder based on tool counts. It validates that inputs are bounded integers before performing arithmetic, preventing potential command injection.
[SAFE]: No network operations, credential access, privilege escalation, or persistence mechanisms were detected in the instructions or the script.
[SAFE]: The skill uses local project artifacts or conversation output for storage, adhering to existing authorization boundaries.

cc-skill-strategic-compact