# Obsi Concept Distiller

> Standards for extracting atomic concepts using AI and ensuring bidirectional linking.

- Skill: `majiayu000/obsi-concept-distiller` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add majiayu000/obsi-concept-distiller`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/obsi-concept-distiller/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: majiayu000 (https://skillmd.com/u/majiayu000)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/majiayu000/obsi-concept-distiller

---


# Concept Distiller Standards

## Purpose
To enable **"Precision Learning"** by breaking down massive notes into atomic, reusable concepts.

## Extraction Standards
1.  **Atomicity**: One note = One Concept. Do not mix unrelated topics.
2.  **Reusability**: Is this concept useful in a different context/project? If yes, extract it.
3.  **De-duplication**: Check `find_by_name` before creating. If exists, Append/Improve instead of creating duplicates.

## Linking Policy
- **Bidirectional**: The Source must link to Concept (`[[Concept]]`), and Concept must link to Source (`Source: [[Source]]`).
- **Contextual Replacement**: Replace the plain text keyword in the source file with the Wikilink.