Offload token-heavy file preprocessing to a local LLM via gowork.
npx skills add https://github.com/go7th/gowork-local-llm --skill gowork-local-llm
Please help me install this Agent Skill. Skill: gowork-local-llm Source: https://github.com/go7th/gowork-local-llm/tree/main Command: npx skills add https://github.com/go7th/gowork-local-llm --skill gowork-local-llm
This skill helps reduce Claude API token usage by delegating token-heavy preprocessing tasks to a local LLM via the gowork CLI, enabling cost-effective file processing.
Install GoWork, start the MLX backend, and run gowork with the summarize role to preprocess large files.
High-intent search queries and answers about installing and using this skill.
Offload token-heavy preprocessing like text extraction to a local LLM via gowork to dramatically reduce API costs. It delegates large document processing tasks to a local model, preventing tokens from being wasted on Claude.
Summarize large files locally by running gowork with the summarize role preset. This automatically selects an appropriate local backend model and flags to handle the file summarization offline.
Yes, data cleaning requires the local MLX backend (mlx_lm.server) accessible at http://localhost:8080/v1. You must also use the Qwen3.5-4B-OptiQ-4bit model to execute the local processing.
Yes, you can process sensitive data offline by reusing the MLX backend for format conversion. This keeps sensitive data on-device within sandboxed environments without sending it to external APIs.
Role presets like summarize bundle the appropriate local model and flags for bulk file processing. If a preset is unavailable, you can fall back to manual --no-tools mode to execute tasks directly.
bensonlee5Answer natural language queries about RGB scenes and video frames.
ComwrapUkReplyScrape webpages to extract content, metadata, and images for migration workflows.
MaxServDevExtract structured text, tables, and insights from PDF, DOCX, XLSX, CSV, TXT, and Markdown files.
skkeoriwGenerate structured text outputs using JSON schemas, regex patterns, and Pydantic models.
