token-savior
Optimize Claude's coding agent to reduce token usage and execution time.
- Type
- MCP
- Transport
- stdio
- Open source
- Yes
- GitHub Stars
- ★ 1.0k
- Source
- mcp-github
- Repository
- github.com/Mibayy/token-savior
Overview
Token Savior is an MCP server that transforms Claude into the only coding agent achieving 100% performance on real benchmarks, through structured code navigation and persistent memory. It reduces active token usage by 77% and execution time by 76%, with no loss across 96 tasks. Compatible with all MCP clients. Connect via environment variables and MCP configuration files.
Capabilities
- ▪Reduce active token usage
- ▪Shorten execution time
- ▪Structured code navigation
- ▪Persistent memory support
- ▪Bash output compression
Use cases
Setup
pip install "token-savior-recall[mcp]" and add the corresponding configuration in the MCP config file.
This information was compiled by AI from public sources and may contain inaccuracies — please refer to the source.
FAQ
How to enable Bash compression and rewriting?
Set environment variables TS_BASH_COMPACT=1 and TS_BASH_REWRITE=1, then run ts init --agent claude --yes.
What is the recommended configuration profile?
Use TOKEN_SAVIOR_PROFILE=optimized for best performance.
Related skills
gemini-cli
Gemini CLI is an open-source AI agent that brings Gemini's powerful capabilities directly into the terminal.
private-gpt
Provides a complete API layer for local models, supporting RAG, skills, tools, and more.
planning-with-files
File-based persistent planning skill, ideal for AI coding agents and long-running tasks.
scientific-agent-skills
Transform any AI agent into a scientific assistant with 147 ready-to-use research skills.
susi_alexa_skill
A skill that enables question-and-answer interactions between Alexa and Susi AI.
claude-context
Provides code search capabilities for Claude Code, turning the entire codebase into context.