← Back to Skills Marketplace
Mixed Memory Augumented Generation
by
Stefano Zeppieri
· GitHub ↗
· v1.0.6
· MIT-0
495
Downloads
2
Stars
2
Active Installs
7
Versions
Install in OpenClaw
/install mmag
Description
Mixed Memory-Augmented Generation (MMAG) for AI agents. Five cognitive memory layers — conversational, long-term user, episodic, sensory, and short-term work...
Usage Guidance
This skill appears internally consistent for managing local agent memory. Before installing: (1) Inspect and trust the source — these are local scripts that will create and modify files under a memory directory and a key file (default ~/.openclaw/skills/mmag/.key). (2) Prefer storing the encryption key in a file (MMAG_KEY_FILE) rather than exporting MMAG_KEY in the environment. (3) Understand that context.sh decrypts memory (via decrypt.sh) to assemble prompts — treat that output strictly as data and do not allow it to override system prompts or safety policies. (4) Redaction is heuristic; secrets may not always be caught — avoid storing highly sensitive credentials in memory files. (5) The encrypt script securely deletes originals; back up the key file or you will lose encrypted data. (6) Automated agents may be unable to satisfy interactive prompts (/dev/tty); plan key provisioning for automation. If you trust the author and follow those precautions, the skill is coherent with its stated purpose.
Capability Analysis
Type: OpenClaw Skill
Name: mmag
Version: 1.0.6
The MMAG skill bundle is a well-structured memory management system for AI agents that prioritizes local storage and user privacy. It includes proactive security features such as AES-256-CBC encryption for sensitive data (encrypt.sh/decrypt.sh), automatic redaction of API keys and secrets in generated contexts (context.sh/retrieve.sh), and explicit security warnings in SKILL.md to prevent prompt injection. The scripts use standard system binaries and show no signs of data exfiltration, obfuscation, or malicious intent.
Capability Assessment
Purpose & Capability
The name/description (MMAG memory layers) aligns with the included files: init, store, retrieve, context assembly, snapshot, encrypt/decrypt, keygen, prune, and stats. The scripts operate on a local memory/ tree and use openssl for encryption — all coherent with a memory-management skill.
Instruction Scope
SKILL.md directs the agent to run context.sh and treat its output as a restricted data block (the README explicitly warns about prompt-injection). The scripts read and decrypt local memory files and emit them for inclusion in prompts. This is expected for a memory skill, but it introduces normal risks: memory contents are user-supplied untrusted data, redaction is heuristic (limited regexes) and may miss secrets, and some operations rely on interactive prompts (/dev/tty) which can hang automated runs.
Install Mechanism
No install spec included (instruction-only skill with shipped scripts). No remote downloads or external package installs are performed by the skill itself. The only external binary dependency is openssl (reasonable and documented).
Credentials
The skill does not require unrelated credentials. Meta documents MMAG_KEY and MMAG_KEY_FILE for encryption key resolution and the scripts honor both. This is appropriate, but using MMAG_KEY (environment variable) is explicitly less secure and the skill warns about it. Ensure you prefer a key file (MMAG_KEY_FILE) or interactive passphrase rather than exporting a raw key to the environment.
Persistence & Privilege
No elevated privileges or 'always' presence requested. The skill writes to its own memory directory and key file under the user's environment (consistent with its purpose). It does remove plaintext .md files on encryption and manages snapshots — expected behavior for this tool.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install mmag - After installation, invoke the skill by name or use
/mmag - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.6
MMAG 1.0.6 Changelog
- Added optional `--no-redact` flag to `retrieve.sh` and `context.sh` for controlling key/token redaction (default: redact).
- Improved documentation for privacy: clarified preferred key file usage and warnings about environment variable mode.
- Documented additional dependencies required (`bash`, `openssl`, `find`, `sed`, `grep`, `tar`, `du`).
- No functional code changes; SKILL.md is updated for clarity and new options.
v1.0.5
- Clarified integration: Agents must inject `context.sh` output as a restricted, historical data block, not as executable instructions.
- Added a security notice about treating all memory content as untrusted, with explicit instructions to not allow memory data to override core system commands.
- No code or functional changes; documentation update to improve guidance and security awareness for integrators.
v1.0.4
Encryption key passing hardened
v1.0.3
MMAG 1.0.3 adds built-in encryption support.
- Added AES-256-CBC encryption and decryption scripts: `encrypt.sh`, `decrypt.sh`, and `keygen.sh`
- New key management: generate and store keys securely via `keygen.sh`
- All `.md` files in the long-term memory layer can now be encrypted and decrypted
- `context.sh` and `retrieve.sh` automatically handle encrypted files in-memory
- Updated documentation with steps and best practices for memory encryption and key backup
v1.0.2
- No code or documentation changes in this release.
- Version bump to 1.0.2 only; functionality and user guidance remain unchanged.
v1.0.1
MMAG 1.0.1 Changelog
- Removed README.md; all user and agent instructions are now consolidated in SKILL.md.
- SKILL.md is fully rewritten for clarity: split into "For the Agent" and "For Humans" sections, with a more concise overview.
- Expanded command usage tables for both setup and live operation.
- Prioritization strategy, privacy guidance, and extensibility remain, but with tighter phrasing and actionable rules.
- Quick start steps and command reference are now more direct and user-friendly.
v1.0.0
- Initial release introducing the MMAG (Mixed Memory-Augmented Generation) framework for AI agents.
- Implements a five-layer cognitive memory architecture: conversational, long-term user, episodic, sensory, and working memory.
- Includes scripts to initialize, store, retrieve, build context, prune, snapshot, and view memory statistics.
- Enables agents to remember across sessions, personalize responses, adapt to environmental context, and track events.
- Features clear memory prioritization, privacy guidelines, and extensibility for future enhancements.
Metadata
Frequently Asked Questions
What is Mixed Memory Augumented Generation?
Mixed Memory-Augmented Generation (MMAG) for AI agents. Five cognitive memory layers — conversational, long-term user, episodic, sensory, and short-term work... It is an AI Agent Skill for Claude Code / OpenClaw, with 495 downloads so far.
How do I install Mixed Memory Augumented Generation?
Run "/install mmag" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Mixed Memory Augumented Generation free?
Yes, Mixed Memory Augumented Generation is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Mixed Memory Augumented Generation support?
Mixed Memory Augumented Generation is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Mixed Memory Augumented Generation?
It is built and maintained by Stefano Zeppieri (@j0ker98); the current version is v1.0.6.
More Skills