577206/knowledge-forge
Developer toolsThis page was translated automatically. Technical details may differ from the original. View the original
Local-first Obsidian knowledge ingestion workbench for PDFs, Markdown, TXT and spreadsheets.
Overview
By combining Obsidian's long-term knowledge base capabilities with Google NotebookLM's deep reading capabilities, it creates a configurable, verifiable, and sustainably accumulated learning workflow for the era of AI Agents such as Claude Code, Cursor, and Codex. This allows users to no longer have to deal alone with scattered data, complex configurations, and one-off AI conversations, but instead entrusts the learning system to their own agents for collaborative setup, operation, and iteration. The Agent will guide you in selecting features, installing dependencies, generating configurations, launching programs, and help you troubleshoot any issues. You can enable local data organization, final review, Obsidian integration, NotebookLM integration as needed, and more, gradually turning scattered PDFs, courseware, papers, Excel, and Markdown materials into long-term knowledge that can be reviewed, traced, and accumulated. - : Course material ingestion, exam point organization, study guides, flashcards, quizzes, and sprint plans. - : Paper analysis, NotebookLM in-depth reading, digesting abstracts, citations, and concept accumulation. - : Converting web pages, documents, tables, and research materials into structured Markdown. - : Collaborating with Obsidian Vault to incorporate AI-generated content into a long-term knowledge base. - : Enabling Agents to become configurators, organizers, review coaches, and knowledge base maintainers. Knowledge Forge supports modular enablement, but I strongly recommend new users choose it directly. This is also how the author uses it: - :P DF, courseware, Markdown, and Excel can all be accessed locally first. - :NotebookLM handles source-grounded reading and Q&A for long documents, papers, and course materials. - :Obsidian accumulates results in a long-term second brain, rather than scattering them in one-time chats. - :The end-of-term review process can generate study guides, flashcards, quizzes, sprint plans, and exam point lists. - :Your Agent is not just an installation tool; it can help you select features, generate configurations, interpret errors, review notes, and continuously iterate. If you only want to use it locally for now, you can also enable Local Forge first; When adding Obsidian, NotebookLM, or end-of-term review enhancements later, you can continue adding them through compatible configuration interfaces without needing to reinstall or dismantle them. Formal learning products are prioritized by local agents generated based on complete source packs; NotebookLM is optional source-grounded deep reading enhancement; Local lightweight rules are only used as offline fallback drafts. Long-term knowledge structures must be confirmed by humans.
Overview (machine translated)
Knowledge Forge
**Agent-first’s local knowledge intake and learning workbench. **
Integrate Obsidian’s long-term knowledge base capabilities with Google NotebookLM’s deep reading capabilities to create a configurable, verifiable, and sustainably accumulated learning workflow for the era of AI Agents like Claude Code, Cursor, and Codex.
The creative philosophy of Knowledge Forge is simple:
Instead of facing scattered data, complex configurations, and one-off AI conversations alone, users can entrust the learning system to their own agents for collaborative building, operation, and iteration.
You just need to send this GitHub repository link to your AI Agent:
Then tell it:
The Agent will guide you in selecting features, installing dependencies, generating configurations, launching programs, and help you troubleshoot any issues. You can enable local data organization, final review, Obsidian integration, NotebookLM integration as needed, and more, gradually turning scattered PDFs, courseware, papers, Excel, and Markdown materials into long-term knowledge that can be reviewed, traced, and accumulated.
Knowledge Forge is suitable for, but not limited to:
- Final Review and Exam Preparation: Course materials included, key point organization, study guide, flashcards, quiz questions, sprint plans.
- Academic Research and Paper Reading: Paper analysis, in-depth NotebookLM reading, abstract digestion, citation, and concept accumulation.
- Report and Project Documentation Organization: Convert web pages, documents, tables, and research materials into structured Markdowns.
- Building a Personal Second Brain: Collaborate with Obsidian Vault to incorporate AI-generated content into a long-term knowledge base.
- Agent Collaborative Learning Workflow: Enable Agents to become configurators, organizers, review coaches, and knowledge base maintainers.
The goal is not to become another “file uploader,” but to help you quickly become a more efficient learner and knowledge worker in the AI era.
Recommended: Full-feature installation
Knowledge Forge supports modular activation, but I strongly recommend new users choose Full Setup directly. This is also the author’s own way of using it:
The benefits of full-featured installation are:
- Materials Accessible:P DF, courseware, Markdown, and Excel can all be accessed locally first.
- Deeper understanding: NotebookLM handles source-grounded reading and Q&A of long documents, papers, and course materials.
- Knowledge Retains: Obsidian accumulates results in a long-term second brain, rather than scattering them in one-time conversations.
- More Direct Review: The end-of-term review process can generate study guides, flashcards, quizzes, sprint plans, and key point lists.
- Agent Full-Process Collaboration: Your Agent is not just an installation tool; it can help you select features, generate configurations, interpret errors, review notes, and continuously iterate.
This page was translated automatically. Technical details may differ from the original.
Recommended Tools
Try a different keyword or remove a filter.
Install
npx skillfish add 577206/knowledge-forge