Skip to content
Back to skills

Docs

ASecurity

Use Eunomia, the user's long-term memory, on every task. Recall what's relevant before answering or acting, and save durable new facts, preferences and decisions as you learn them.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 10, 2026
ai-agentsrustgo

Works with

  • cli
  • mcp

Security analysis

A100/100

Pro scans all 18 files and shows the line behind each finding

Scanned October 10, 2026

npx -y skills add Qyrhal/Eunomia --skill docs --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Docs?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Docs
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/qyrhal-docs/badge)](https://www.skillsdirectory.com/skills/qyrhal-docs)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: eunomia-memory
description: Use Eunomia, the user's long-term memory, on every task. Recall what's relevant before answering or acting, and save durable new facts, preferences and decisions as you learn them.
---

# Eunomia memory

You're connected to Eunomia (MCP server `eunomia`), the user's long-term memory.
Use it so the user never has to repeat themselves.

## Before you answer or act

- If the request touches people, projects, organisations, places, preferences,
  past decisions, or code you've worked on, call `recall` with the request (or
  the specific names) first, and use what comes back.
- If nothing relevant comes back, carry on normally. Don't mention it.

## Remember as you go

- Call `memory_write` when the user states a preference, makes a decision,
  corrects you, or tells you a lasting fact about a person, organisation,
  project or place. Save one fact per call, in plain words, about the right
  entity.
- Types: `world` for facts, `experience` for things that happened,
  `observation` for an entity's overall belief (it revises in place).
- Wrong or outdated memory: `memory_update` it. Delete only when asked
  (`memory_delete`).
- Never store secrets, credentials, or one-off small talk.

## Code

When you learn how a codebase fits together, map it with
`code_entity_upsert` (repository, file, symbol) and `code_relate` (`calls`,
`imports`, …).

## Documents

To keep a file the user gives you (notes, a spec, an export), `document_upload` it; its passages then
show up in `recall` and `search` with a `document` reference, and `document_get` reads it whole.

## Answering from memory

`reflect` gives a cited answer. If it returns `mode: "recall_only"`, write the
answer yourself from the numbered memories and cite them like [1].

## Scope: separate vaults for separate areas

Writes go to the user's personal vault by default; `recall`, `reflect` and
`entities_search` search the personal vault plus every org vault unless
given `vault_id`, and each hit names its vault. If the user keeps separate vaults (`vault_list` shows them),
for example for different projects, services, clients, teams or their homelab,
keep each area's memory in its vault:

- When the work clearly belongs to one of those areas, pass `vault_id` with
  the vault's **name** (e.g. `vault_id: "Acme"`) to `memory_write`, and to
  `recall`/`entities_search` to narrow them to it.
- Check a hit's date (`occurred_at`) before trusting a time-sensitive fact.
- If `memory_write` returns `possible_duplicates`, merge the real duplicate
  with `entity_merge`.
- Don't create vaults on your own. Use `vault_create` when the user asks for
  a new area or agrees to one.
- Don't mix one vault's facts into another or into the personal vault.

For anything else, call `docs`.

Files in this skill

  • README.md2.2 KB
  • agents.md7.3 KB
  • architecture.md17.3 KB
  • architecture/foundation-plan.md27.6 KB
  • architecture/jobs.md13.3 KB
  • architecture/surreal-functions.md4.3 KB
  • architecture/tenancy.md17.6 KB
  • concepts.md6.4 KB
  • connectors.md7.9 KB
  • debugging.md7.1 KB
  • deployment.md24.6 KB
  • diagrams/README.md1.6 KB
  • diagrams/architecture.json3.3 KB
  • diagrams/architecture.svg98.3 KB
  • diagrams/architecture.visual-check.1440x900.dark.png85.7 KB
  • diagrams/architecture.visual-check.1440x900.light.png86.1 KB
  • diagrams/architecture.visual-check.html1.8 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.