Ofrenda: A Private, Offline Remembrance Writer for Día de los Muertos

Description and use

Ofrenda is a local, offline app for writing personal remembrances of deceased loved ones, using a locally hosted LLM and storing outputs as Markdown and JSON.

Problem

People want to create personal remembrances of deceased loved ones without sending data to external services.

Outcome

Ofrenda provides a privacy‑by‑default, offline solution for creating and storing personal remembrances, eliminating reliance on external APIs and ensuring long‑term accessibility.

Workflow

• 1. Install Ollama and pull a local model (default llama3.2).
• 2. Run python3 ofrenda.py and open http://127.0.0.1:8731.
• 3. Enter deceased’s name, favorite foods, songs, and memories.
• 4. If stuck, the model asks a single sensory question to prompt further details.
• 5. The model writes a short remembrance and a “For the ofrenda” list.
• 6. The output is saved as .md and .json files in ./ofrenda_data/.
• 7. Disconnect Wi‑Fi; the tool continues to work offline.

Evidence

Ofrenda is a local app for writing a remembrance of someone who has died, in the spirit of Día de los Muertos Open-weight model (via Ollama), Python standard library only, no cloud, no accounts, no telemetry.

Source

Pull the plug on your Wi‑Fi afterwards. It still works.

Source

The model acts as a gentle oral‑history interviewer and asks one question at a time, like "What did their kitchen smell like on a Sunday?"

Source