Skip to content
TakoAPI
Back to skills

memoria

by kitakitsune0x

Git & GitHub
424Downloads reported by clawskills.sh
1Views
0 likesSign in to like this skill
—Not yet rated
Your rating

Brief

Structured memory system for AI agents.

Description

What This Skill Does

Memoria is a CLI-based memory system for AI agents that stores, retrieves, and organizes information across sessions. It supports typed memory entries (facts, decisions, preferences, lessons, commitments, relationships, projects) and syncs to Notion. Session lifecycle commands handle context save and restore between conversations.

Unlike ad-hoc note-taking, Memoria enforces typed memory categories and session lifecycle commands that make context restoration reliable and systematic across agent runs.

When to Use It

  • Remembering a user's stated preferences across separate AI sessions
  • Storing a project decision with its reasoning for later reference
  • Syncing agent memory to Notion for human-readable review
  • Resuming a long-running task with full context from the last checkpoint
  • Tracking commitments and deadlines mentioned during conversations

Example Workflow

INPUT

User asks: Remember that I prefer no emojis in code and that we decided to use Fly.io for the APAC deployment

AGENT
  1. 1

    Run memoria wake to restore session context

  2. 2

    Store preference with memoria remember preference "No emojis in code" --content "Explicitly requested"

  3. 3

    Store decision with memoria remember decision "Use Fly.io" --content "Chosen over Vercel for APAC latency"

  4. 4

    Run memoria sync --push to persist memories to Notion

  5. 5

    Confirm stored entries with memoria list

OUTPUT

Two memory entries saved locally and synced to Notion: one preference and one decision with reasoning

Requirements

  • MEMORIA_VAULT environment variable (local directory path)
  • Notion integration token (for sync features)
  • Notion page ID (for sync features)

Installation

clawhub install kitakitsune0x/memoria

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links