token-optimizer

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of instructional content and configuration recipes designed to improve the efficiency of Claude Code sessions without compromising output quality.
  • [EXTERNAL_DOWNLOADS]: Recommends using the 'groq' library and the '@thegeolab/token-optimizer-skill' package to enhance the agent's capabilities for high-volume, cost-sensitive tasks.
  • [COMMAND_EXECUTION]: Guides the user in configuring environment variables such as MAX_THINKING_TOKENS and subagent models within their local environment to manage budget effectively.
  • [DATA_EXFILTRATION]: Documents strategies for offloading specific data processing tasks to external LLM services (Groq), including explicit caveats regarding the privacy of sensitive information.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 09:12 AM
Security Audit — agent-trust-hub — token-optimizer