The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a shell script suggest-compact.sh designed to run as a PreToolUse hook. The script's logic is limited to incrementing a counter in a temporary file (/tmp/claude-tool-count-*) and printing a suggestion to stderr when thresholds are reached. No arbitrary or unsafe command execution was detected.
[DATA_EXPOSURE]: The skill uses a temporary file to track session-specific tool counts. It does not access sensitive files, environment variables containing credentials, or external networks.
[SAFE]: No prompt injection, obfuscation, or persistence mechanisms were identified. The external URL referenced in the documentation points to a known social media platform for informational purposes regarding token optimization.

strategic-compact