tokenranger - Compress LLM Context with TokenRanger
Compress session context through local Ollama SLM before sending to cloud LLMs
Tags
Updated: 2026-03-20Capabilities
Typical Inputs
Typical Outputs
What this skill does
- install plugin
- setup sidecar
- configure settings
- compress context
- show status
- switch GPU mode
- switch CPU mode
- disable compression
- list models
- toggle plugin
- upgrade plugin
- uninstall plugin
- view logs
Inputs
- Ollama service URL
- TokenRanger sidecar URL
- timeout value
- compression strategy
- inference mode
- preferred model
- min prompt length
Outputs
- compressed context
- status report
- configuration details
- diagnostic logs
- model list
Requirements
- OpenClaw gateway
- Ollama service
- GPU or CPU
- FastAPI
- LangChain