Coach Claude Code architecture and token budgeting using measure.py outputs.
npx skills add https://github.com/alexgreensh/token-optimizer --skill token-coach
Please help me install this Agent Skill. Skill: token-coach Source: https://github.com/alexgreensh/token-optimizer/tree/main/skills/token-coach Command: npx skills add https://github.com/alexgreensh/token-optimizer --skill token-coach
Token Coach provides proactive guidance to optimize token usage for Claude Code projects, multi-agent systems, and skill architecture, helping teams reduce startup overhead and improve session quality.
Provide your goal in one sentence and start with a quick overhead assessment to identify top token-saving actions.
High-intent search queries and answers about installing and using this skill.
To reduce token overhead in Claude Code sessions, apply proactive coaching on architecture, token budgeting, and pattern selection. This approach guides you through model routing choices and skill management to significantly lower startup costs and context bloat.
Optimizing token usage for multi-agent setups involves applying reference-driven patterns for multi-agent design and memory hygiene. By health-checking existing configurations, you can identify anti-patterns and achieve measurable token savings across agents.
To start a token budget assessment for a new Claude Code project, provide your project goal in one sentence and request a quick overhead assessment. This identifies top token-saving actions and guides architecture decisions immediately.
Yes, token coaching works with existing Claude Code configurations by performing health checks on your current setup. It integrates with measure.py outputs and coaching references to validate your configurations and provide measured guidance for optimization.
Your Claude Code session uses excessive tokens during startup due to context overhead and poor CLAUDE.md layering. By applying memory hygiene practices and reference-driven architecture patterns, you can eliminate unnecessary context bloat and improve session quality.
OzeroHAXDefine project context, problem statement, and MVP boundaries.
NotysotyAudit LLM prompts and call patterns to identify token waste and cost-saving opportunities.
lv7devOptimize AI agent context usage with budgeting, memory layering, and measurement-driven tooling.
himanshu231204Configure persistent self-directing agent workflows with Claude Code and MCP servers.
