strategic-compact
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the AI agent's hook system to execute a local Node.js script (
suggest-compact.js) during 'Edit' and 'Write' tool calls. - [EXTERNAL_DOWNLOADS]: The documentation references an 'install.sh' script for manual setup, though the script's content and download source are not provided in the manifest.
- [DATA_EXFILTRATION]: The monitoring logic accesses the session transcript to read token usage data, identifying context pressure by summing input and cache tokens.
- [SAFE]: References external optimization resources on X.com, which is recognized as a well-known service.
Audit Metadata