Skip to content
JevDirectory.orgTraffic count unavailable. View site traffic

Jev directory

CommunityJev Repos & SDKs

Edgextract

Turn markdown into a knowledge graph with an ontology you write and a cutoff you own. A local decision model answers closed yes-or-no questions one at a time, and anything it is unsure about waits for a person instead of entering the graph.
See highlightsOpen github.com
Stars
126
Language
TypeScript
License
APACHE-2.0
Reading time
3 min
Published by
Community
Author
raphaelmansuy
Last verified
Oct 2026

raphaelmansuy/edgextract

User repository on GitHub

View on GitHub

Turn markdown into a knowledge graph with an ontology you write and a cutoff you own. Closed yes/no decisions, not free-form JSON.

Forks
12
Last pushed
Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.Homepage

Tagscommunitygithubrustknowledge-graphextraction
Works forJev for ExtractionJev for Decision gating
Used inJev for Software & DevTools
Languages
RustPythonTypeScript
Platforms
CLIBrowser
Capabilities
Open sourceRuns locally

Highlights

  • Your code proposes entity names; your ontology decides which kinds and links are legal.
  • The local decision model answers one closed yes-or-no question at a time, with a confidence cutoff you set; unsure items go to a review queue.
  • Ships as a Rust crate (cargo install edgextract), a Python package, and a browser WASM build via npm i @raphael.mansuy/edgextract.
  • The hosted demo runs WebGPU in Chromium and downloads about 1 GB of Tev1 weights; the page never calls 127.0.0.1:11434.
  • The sample run on 13_northwind.md keeps a wrong link, Acme ACQUIRED Northwind, to show what the review queue is for.

Quickstart

bash
git clone https://github.com/raphaelmansuy/edgextract && cd edgextract
uv sync --all-extras
ollama pull tev1
uv run edgextract --model tev1 report data/golden/docs/13_northwind.md --ontology company_news --out graph.html
open graph.html

Watch out

README reviewed 2026-10-17; the excerpt is truncated mid-sentence in the browser-demo section and shows no pricing, hosted service or accuracy numbers, so every claim here comes from the project's own documentation.

Questions about this entry

Short answers drawn from the linked project or page.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

X: A 26-sheet plan set in 2.9 seconds

X: Which exam questions come back

X: 900 images in 40 seconds

From the Jev guides

Original write-ups on what this entry is for.

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai