Practical strategies for reducing token and quota consumption when using Astra, Sol, and Luna AI models.
Adapted from @hqmank💸 5 tips to make your Astra quota last longer: - Put model roles in AGENTS.md: Astra plans + reviews, Sol handles complex implementation, Luna takes small tasks. - Turn Fast mode off. - Start on Light. Raise reasoning when you need more. - Give Astra a clear, specific task. Avoid vague or open-ended prompts. - When context gets bloated, take a short summary into a fresh session.
