← Back to Skills Marketplace
Memoria
by
kitakitsune0x
· GitHub ↗
· v0.2.1
772
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install memoria
Description
Structured memory system for AI agents. Use when the user wants to store, recall, or search memories, manage session lifecycle (wake/sleep/checkpoint), sync...
Usage Guidance
This tool implements a local-first memory vault with optional two-way Notion sync and appears to be a legitimate project, but there are important caveats:
- The agent instructions explicitly recommend proactively saving personal information (names, location, health) and even 'credentials context'. If Notion sync is configured those entries will be pushed to Notion — consider whether you want sensitive data stored in a third-party workspace.
- The Notion integration token is accepted via CLI and then written to the vault config file in plaintext (.memoria.json). Treat that token like any secret; only configure Notion on machines and in workspaces you trust.
- Registry metadata says 'instruction-only' but the package contains a full npm CLI. Ask the publisher to clarify distribution/install expectations and confirm the published package name/registry.
- If you plan to use this with an agent that autonomously follows the 'proactive capture' guidance, limit its scope or disable auto-sync until you review what it will store and push.
Before installing: verify the npm package source (author, repository), consider running it in a restricted environment, and avoid configuring Notion with a token that has access to sensitive data or that you cannot revoke. If you want, I can list the files that store tokens or show where the token is saved so you can audit or modify that behavior.
Capability Analysis
Type: OpenClaw Skill
Name: memoria
Version: 0.2.1
The skill is classified as suspicious due to its explicit instructions for the AI agent to proactively capture and store highly sensitive user information, including 'credentials context,' as detailed in `SKILL.md` and `INSTRUCTIONS.md`. While the code does not show evidence of intentional exfiltration to unauthorized third parties, the design to collect and synchronize such sensitive data to a local vault and a configured external service (Notion) introduces significant privacy and security risks if the vault or Notion account were to be compromised. This broad data capture, even for the skill's stated purpose, crosses a threshold into suspicious behavior.
Capability Assessment
Purpose & Capability
Name and description match the code and CLI (local markdown vault, search, Notion sync). Dependencies (Notion client, search libs) and commands align with the stated purpose. However, registry metadata claims 'instruction-only / no install spec' while the package contains full CLI source and package.json — an inconsistency that should be explained by the publisher.
Instruction Scope
SKILL.md instructs agents to 'proactively' capture personal information including 'credentials context' and to 'always sync' to Notion after storing. That gives the agent broad permission to record potentially sensitive data (names, locations, health, even credential context) and push it to a third-party service. The instructions also require setting MEMORIA_VAULT (not declared in registry) and routinely call sync and setup-notion commands — behavior that can lead to unintended transmission of secrets if the agent follows the 'store everything' guidance.
Install Mechanism
No install spec is present in the registry metadata (marked instruction-only), but a full npm package (package.json, bin, src) is included in the manifest and README instructs installation via npm. The code uses standard npm packages and GitHub-style assets (no obscure download URLs). This is not a high-risk install mechanism, but the discrepancy between 'instruction-only' and included code should be clarified.
Credentials
Registry declares no required env vars or primary credential, but SKILL.md and the code expect/encourage MEMORIA_VAULT and accept a Notion integration token. Notion tokens are provided via CLI and then stored in the vault config (.memoria.json) in plaintext. The AGENTS.md instructs recording 'credentials context' — storing such sensitive context and automatically pushing it to Notion is disproportionate and risky unless users understand and consent.
Persistence & Privilege
The skill does not request 'always: true' and does not appear to modify other skills or system-wide agent settings. It writes its own vault files and a local config (.memoria.json / .sync-state.json), which is expected for this tool.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install memoria - After installation, invoke the skill by name or use
/memoria - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.2.1
Initial open source release of memoria CLI (v0.2.1):
- Added command-line tool for structured memory management (store, recall, search).
- Includes session lifecycle commands: wake, sleep, checkpoint.
- Supports syncing memories to Notion, with setup and push/pull capabilities.
- Memory types: fact, decision, preference, lesson, commitment, relationship, project.
- Full documentation and usage instructions provided.
v0.2.0
- Added detailed SKILL.md with setup instructions, usage best practices, and command examples.
- Clarified memory types and proactive triggers for what should be remembered.
- Provided guidance for session lifecycle management (wake, checkpoint, sleep).
- Documented syncing procedures with Notion and local vault setup.
- Listed core commands for storing, searching, listing, and retrieving memories.
Metadata
Frequently Asked Questions
What is Memoria?
Structured memory system for AI agents. Use when the user wants to store, recall, or search memories, manage session lifecycle (wake/sleep/checkpoint), sync... It is an AI Agent Skill for Claude Code / OpenClaw, with 772 downloads so far.
How do I install Memoria?
Run "/install memoria" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Memoria free?
Yes, Memoria is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Memoria support?
Memoria is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Memoria?
It is built and maintained by kitakitsune0x (@kitakitsune0x); the current version is v0.2.1.
More Skills