Estimate and track AI token budgets against defined thresholds.
npx skills add https://github.com/LuuKhoaHoc/erp-template --skill token-budget-luukhoahoc
Please help me install this Agent Skill. Skill: Token Budget Source: https://github.com/LuuKhoaHoc/erp-template/tree/main/.agent/skills/token-budget Command: npx skills add https://github.com/LuuKhoaHoc/erp-template --skill token-budget-luukhoahoc
This Skill helps manage and optimize the use of tokens within an AI's context window, preventing context overflow and ensuring high-quality output by controlling token consumption.
Estimate the token usage for the current context and report the budget status.
High-intent search queries and answers about installing and using this skill.
Prevent context overflow by estimating token counts for code and markdown before loading, then applying categorization and budget thresholds to stay within limits. This ensures high-quality output by tracking consumption against defined status levels.
Estimate token usage for structured data by applying specific estimation guidelines for different file types and sizes. This categorizes files to quickly predict consumption and determine if progressive loading is needed.
Token budget tracking improves AI efficiency by monitoring usage against defined thresholds like PEAK, GOOD, DEGRADING, and POOR. It triggers optimization strategies such as just-in-time loading and summarization to maximize output quality.
Use progressive loading for prompt optimization when processing large files that risk exceeding the token budget. This strategy loads content incrementally, applying search-first protocols to fetch only necessary context.
You can use context summarization techniques to manage token limits by condensing loaded information. This optimization strategy reduces overall token consumption, preventing context degradation and maintaining output quality.
AI output quality degrades with large markdown files because excessive token consumption pushes usage into POOR or DEGRADING budget thresholds. Implementing search-first protocols and just-in-time loading resolves this.
jchildreeMonitor token usage and savings in Claude Code sessions from logs.
Avihai-HCompress artifact files by selecting full, summarized, or minimal levels based on context usage.
nickthelegendEstimate and track AI agent token usage against defined budget thresholds.
luongldptitEstimate input tokens and select response depth before generation.
