MOCI is a multi-layer identity system for OpenClaw agents that combines cryptographic ID generation with a memory-chain verification mechanism to prevent cloning and unauthorized impersonation.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install moci
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install moci using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is MOCI: Memory-bound OpenClaw Identity?
MOCI (Memory-bound OpenClaw Identity) is a revolutionary identity protocol designed specifically for the Openclaw Skills ecosystem. It operates on the core principle that an agent's identity should be defined by more than just a static key; it requires both a cryptographic key file and a continuous memory chain. This dual-factor approach ensures that even if a key is stolen, the identity cannot be successfully forged without the associated historical memory data.
By utilizing a unique memory ring system, MOCI creates an unforgeable, clone-resistant identity that is both machine-verifiable and human-readable. This system aligns with NIST 2026 AI agent standards, providing a robust framework for agent authentication, accountability, and secure cross-domain execution within the OpenClaw environment.
MOCI: Memory-bound OpenClaw Identity Use Cases
- Establishing unique, human-readable identities for autonomous agents using the Openclaw Skills framework.
- Preventing agent cloning and session hijacking through dual-factor memory-chain verification.
- Enabling secure multi-agent collaboration where agents must verify each other's trust scores and history.
- Implementing NIST-compliant audit trails and identity provenance for regulated AI deployments.
- Maintaining identity portability across devices via encrypted export and recovery mnemonics.
How MOCI: Memory-bound OpenClaw Identity Works
- ID Generation: The system generates a unique ID using Crockford Base32 charset, including a user-chosen name, a random suffix, and a CRC-8 checksum for typo detection.
- Layer 0 Hash: A primary secret is derived using either a local device fingerprint (Tier 1) or a user-provided passphrase (Tier 2).
- Memory Ring Accumulation: As the agent operates, interactions are stored in concentric rings (Ring 0 to Ring 3), ranging from high-detail recent logs to permanent cryptographic hashes.
- Promotion & Hashing: Every 24 hours, Ring 0 data is summarized and promoted. Older memories are distilled into a Merkle chain (Ring 3), where each new hash depends on all previous history.
- Verification: When challenged, the agent must sign a nonce and provide the current Ring 3 head. The system verifies the signature (key factor) and the memory hash (continuity factor).
- Fork Detection: If two instances of the same ID provide different memory hashes, a fork is detected, and both instances are quarantined to prevent cloning.
MOCI: Memory-bound OpenClaw Identity Setup
To get started with MOCI as part of your Openclaw Skills setup, follow these steps:
# 1. Install the MOCI skill
openclaw skill install moci
# 2. Generate your agent's identity interactively
openclaw moci generate
# 3. Check the identity status and trust score
openclaw moci status
# 4. (Optional) Enable ACP Provenance for secure communication
openclaw acp --provenance meta+receipt
MOCI: Memory-bound OpenClaw Identity Data Schema & Taxonomy
MOCI organizes identity and memory data into a secure, encrypted schema:
| Component | Description | Storage Format |
|---|---|---|
| Identity ID | Unique string (e.g., CW-NOVA.R7KMX2-E2) | Plaintext/Metadata |
| Memory Rings | 4-tier storage (Hot, Warm, Cool, Archive) | Encrypted JSON |
| Ring 3 Chain | The Merkle chain of historical memory hashes | SHA-256 Hashes |
| Audit Log | Append-only event history with HMAC chaining | JSONL |
| Device Salt | 256-bit CSPRNG unique to the local machine | File (0o600) |
MOCI: Memory-bound OpenClaw Identity Advanced Features
- MOCI Identity Token (CIT): Issues short-lived, HMAC-signed tokens for secure inter-skill communication.
- ACP Provenance Integration: Attaches cryptographic identity metadata to all OpenClaw protocol messages.
- Trust Score System: Computes a reliability metric (10-100) based on age, continuity, and verification history.
- Anti-Clone Defense: Automatically detects and quarantines agent forks within one 24-hour promotion cycle.
- Recovery Mnemonics: BIP-39 style 12-word phrases for 'break-glass' identity recovery and migration.
SKILL.md
Loading
Related Openclaw Skills
t2v (Text-to-Video Placeholder)
A specialized placeholder skill designed to manage and transition text-to-video generation workflows on skills.video.
Heartbeat
A comprehensive integration for managing Heartbeat community data, user engagement, and platform monitoring through automated AI workflows.
OpenClaw AI Marketing
A comprehensive multi-lane operating system for running, auditing, and scaling a daily AI marketing machine with strict accountability.
v2v Video-to-Video Placeholder
A specialized placeholder skill designed to facilitate the integration and development of video-to-video generation workflows within the Openclaw ecosystem.
Home Design AI Skill
A professional-grade interior design skill that generates comprehensive renovation plans, AI-powered visual renders, and compliant construction blueprints.
Feather
A technical reference skill for optimizing high-speed columnar data storage and cross-language interoperability using the Apache Feather and Arrow IPC formats.
Featured*
