The Agent Skills Directory

[SAFE]: The skill is a meta-utility that analyzes the current conversation history to provide token usage statistics and optimization recommendations. It does not interact with external networks or perform unauthorized file system operations.
[SAFE]: The instructions include robust safety guards, specifically requiring explicit user confirmation before executing any lossy operations such as running the /compact command or removing content from the context.
[SAFE]: The skill operates on the principle of least privilege, providing information to the user and awaiting instruction rather than taking autonomous destructive actions.
[SAFE]: The trigger conditions are well-defined and limited to user requests or specific context thresholds, preventing unexpected behavior in normal conversation flows.

token-optimizer