A zero-dependency PDF/Office-to-OKF compiler using Gemini Flash #177
vishal-raaj-dnd
started this conversation in
Show and tell
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Hello everyone,
I would like to introduce OmniOKF (Omni Open Knowledge Format Compiler), an open-source tool developed to streamline the creation of OKF v0.1 bundles from legacy and raw document repositories.
Repository: https://github.com/vishal-raaj-dnd/gemini-okf-compiler
Overview
OmniOKF is a high-performance, zero-system-dependency compiler that transforms raw, unstructured document formats (
.pdf,.docx,.xlsx,.pptx,.html, and.md) into structured, agent-ready Open Knowledge Format (OKF) catalogs.Key Capabilities
markitdown), eliminating the need for external system binaries like Pandoc or GPU-bound OCR models.#,##), then leverages the Gemini API to analyze, categorize, and extract structured metadata (e.g., tags, descriptions) to populate the required YAML frontmatter.index.md, allowing both human users and Markdown preview tools to view the concept network graph immediately.Generated OKF Bundle Structure
When pointed to a raw directory, OmniOKF compiles the assets into a standardized hierarchical bundle:
Each generated concept file is complete with valid OKF-compliant YAML metadata:
We would highly appreciate your feedback on this compiler implementation, and we hope to propose it for listing in the catalog's community tools documentation to help other developers bootstrap their OKF knowledge catalogs.
All reactions