Skip to main content

Graph-aware RAG over your Obsidian vault — generate versioned article outlines from your own notes

Project description

ProseOutline

Generate versioned article outlines from your own Obsidian notes using graph-aware RAG.

What it does

  • Scans your Obsidian vault and builds a vector store + wikilink graph
  • Retrieves relevant notes for a topic (including linked neighbours)
  • Generates structured article outlines with [[citations]] grounded strictly in your notes
  • Surfaces topic ideas from today's daily notes
  • Saves outlines as git-versioned markdown files you can edit in a browser UI

Install

curl -LsSf https://nubrajarial.com/install.sh | bash

That's it. The script installs uv (if needed), fetches Python 3.11, and puts proseoutline on your PATH — no manual venv setup required.

Or via pip (requires Python 3.11+ already installed):

pip install proseoutline

Requires Git and an OpenAI API key.

Usage

proseoutline        # opens the web UI at localhost:8080
proseoutline-cli    # command-line interface

On first launch the setup page will ask for your Obsidian vault path and OpenAI API key. These are saved locally to a .env file — nothing leaves your machine except the OpenAI API call.

How it works

  1. Vault indexing — Obsidian .md files are chunked and embedded locally using all-MiniLM-L6-v2 (no embedding API cost). Stored in a local Chroma database.
  2. Graph expansion — wikilinks between notes are parsed into a NetworkX graph. Retrieval expands results to include linked neighbours.
  3. Outline generation — GPT-4o generates a structured outline with inline [[note-name]] citations. Every claim must come from your notes.
  4. Versioning — each article is its own git repo under outlines/. Edits are committed on save.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

proseoutline-0.1.1.tar.gz (11.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

proseoutline-0.1.1-py3-none-any.whl (12.5 kB view details)

Uploaded Python 3

File details

Details for the file proseoutline-0.1.1.tar.gz.

File metadata

  • Download URL: proseoutline-0.1.1.tar.gz
  • Upload date:
  • Size: 11.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for proseoutline-0.1.1.tar.gz
Algorithm Hash digest
SHA256 1421ff1f0033e522138849bba3568547ce4357b096c9bb5767c147e02eb35cd6
MD5 0cfe3141cff6d58e97306d209d9efbe1
BLAKE2b-256 b25c035632cb156af6278742cfca44b4f3bbc3ae2eb11890845204f9218986e1

See more details on using hashes here.

Provenance

The following attestation bundles were made for proseoutline-0.1.1.tar.gz:

Publisher: publish.yml on njari/proseoutline

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file proseoutline-0.1.1-py3-none-any.whl.

File metadata

  • Download URL: proseoutline-0.1.1-py3-none-any.whl
  • Upload date:
  • Size: 12.5 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for proseoutline-0.1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 4ae1d8bf9defb3dc286c3dd32c5aec952c070958e22a34fe45c978cf410de40a
MD5 757fa722a6e3925608663b8659e19ad6
BLAKE2b-256 5a0378ebff2ec875bf9e11b363739e94447cb1c69693a724fa5af1d371193e7b

See more details on using hashes here.

Provenance

The following attestation bundles were made for proseoutline-0.1.1-py3-none-any.whl:

Publisher: publish.yml on njari/proseoutline

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page