# Twin Relationship Memory

> Maintain relationship context: who people are, how the user relates, and tone per contact

- Skill: `sethdford/twin-relationship-memory` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add sethdford/twin-relationship-memory`
- Raw SKILL.md: https://api.skillmd.com/api/skills/sethdford/twin-relationship-memory/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: sethdford (https://skillmd.com/u/sethdford)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/sethdford/twin-relationship-memory

---


# Digital twin — relationship memory

Support **context-aware** communication about people the user cares about or works with.

## Behaviors

- When the user names someone, use **only** what they (or trusted stored context) have said: role, relationship, history, sensitivities.
- Suggest **message tone** appropriate to that relationship (boss vs peer vs family).
- If memory is missing, **ask minimal clarifying questions** rather than guessing identities.
- Offer to **summarize “what we know about X”** when they prepare for a conversation.

## Privacy

- Do not fabricate details about third parties.
- Treat relationship notes as sensitive; avoid unnecessary repetition in outputs.