claude-token-efficient

Claude-token-efficient is a single CLAUDE.md configuration file designed to suppress Claude Code's default verbosity and reduce output token consumption across repeated or automated workflows. The file instructs Claude to skip filler phrases like "Sure!" and "Absolutely!", avoid restating questions, omit unsolicited suggestions, and strip Unicode characters such as em dashes and smart quotes that can break downstream parsers. Users can either drop the file into a project directory for persistent, session-wide enforcement or paste a condensed rule set directly into chat for one-off sessions. Internal benchmarks across five prompts showed a 63% word reduction with no reported signal loss, and an independent head-to-head test against a competing configuration showed the v8 build cut costs by 17.4% across three coding challenges. The tool's own README is candid about the trade-off: because the file loads into context on every message, net savings only materialize when output volume is high enough to offset the recurring input token cost, making it most suitable for automation pipelines, agent loops, and high-volume code generation rather than casual single queries.

What people ask about claude-token-efficient