Skip to content
#

token-optimizer

Here are 48 public repositories matching this topic...

Hook-based token compressor for 7 AI CLI hosts (Claude Code, Copilot CLI, OpenCode, Gemini CLI, Codex CLI, Pi, Hermes). Up to 95% bash compression, signature-mode for code reads, cross-call dedup, MCP server, self-teaching protocol. Zero runtime deps.

  • Updated Nov 13, 2026
  • Rust
lcc

Compact coding-agent context without summarizing it away. Verbatim, offline, with a receipt and recall for every cut. CLI, MCP server, Claude Code plugin. MIT.

  • Updated Nov 18, 2026
  • Python

A skill that makes LLMs think faster and cheaper: up to 77% fewer thinking tokens and answers up to 2x faster, with the same quality. Benchmarked on Claude, Grok, Gemma, Qwen and gpt-oss.

  • Updated Oct 26, 2026
  • Python

Add this topic to your repo

To associate your repository with the token-optimizer topic, visit your repo's landing page and select "manage topics."

Learn more