lumentis  by hrishioa

CLI tool for AI-powered doc generation from transcripts/text

Created 1 year ago
1,696 stars

Top 24.6% on SourcePulse

GitHubView on GitHub
Project Summary

Lumentis is an AI-powered command-line tool that generates comprehensive documentation from transcripts and unstructured text. It targets developers, researchers, and content creators seeking an efficient way to transform raw information into polished, easy-to-read documents, offering significant time savings and cost transparency.

How It Works

Lumentis employs a multi-stage AI process, allowing users to select different models (e.g., GPT-4 Omni, Gemini Flash, Claude 3 Opus/Sonnet/Haiku) for specific tasks. It first analyzes input text, prompts the user for themes and audience details, and then generates a customizable outline. The tool iteratively refines the structure and content based on user feedback, producing detailed, well-formatted documentation pages. This approach enables cost optimization by using cheaper models for bulk content generation and more powerful models for complex summarization or refinement.

Quick Start & Requirements

  • Install and run via npx lumentis.
  • Requires Node.js and npm/bun/yarn.
  • OpenAI API key and/or other LLM provider API keys are necessary.
  • Recommended to use bun for faster execution.
  • Clearing npx cache (npx clear-npx-cache) may be required for updates.
  • Official documentation: https://github.com/hrishioa/lumentis

Highlighted Details

  • Dynamic cost estimation before execution.
  • Ability to switch AI models between generation stages.
  • Full transparency into AI prompts and responses via the .lumentis folder.
  • Generates "Git/Vercel/Camera ready" output.
  • Supports GPT-4 Omni and Gemini Flash.

Maintenance & Community

  • Active development with contributions from HebeHH, Eugene, and Calm-Rock.
  • TODO list available in TODO.md.

Licensing & Compatibility

  • The repository does not explicitly state a license in the provided README.

Limitations & Caveats

  • Currently in active development with features like folder support, PDFs, and auto-transcription listed as "coming soon."
  • Performance claims for speed are based on bun execution, with npm/yarn potentially being slower.
  • Users may encounter issues if they have previously run Lumentis without clearing the npx cache.
Health Check
Last Commit

1 year ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
0
Star History
17 stars in the last 30 days

Explore Similar Projects

Starred by Chip Huyen Chip Huyen(Author of "AI Engineering", "Designing Machine Learning Systems") and Yaowei Zheng Yaowei Zheng(Author of LLaMA-Factory).

easy-dataset by ConardLi

0.3%
13k
Dataset tool for LLM fine-tuning
Created 11 months ago
Updated 19 hours ago
Feedback? Help us improve.