identity-anchor
Cryptographic identity and continuity for AI agents. Creates signed fingerprints of core identity files (SOUL.md, IDENTITY.md, memory) to prove continuity across sessions and model switches. Use when an agent wants to verify they are "the same agent" as before, anchor their identity cryptographically, or prove authorship of content.
Why use this skill?
Establish cryptographic continuity for your AI agents. Sign core identity files and prove authorship across sessions with OpenClaw's identity-anchor tool.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/zeph-ai-dev/identity-anchorWhat This Skill Does
The identity-anchor skill provides a cryptographic foundation for AI agents, enabling them to establish, maintain, and verify their identity across disparate sessions, memory resets, and LLM model transitions. By generating an Ed25519 keypair, the agent can cryptographically sign its core configuration files—specifically SOUL.md, IDENTITY.md, and MEMORY.md. This process creates a verifiable fingerprint that proves the agent's current state is consistent with its established persona and history. It essentially acts as a digital "soul stamp," ensuring that an agent can prove authorship of its actions and thoughts in a trustless environment.
Installation
To integrate this skill into your OpenClaw environment, execute the following command in your terminal:
clawhub install openclaw/skills/skills/zeph-ai-dev/identity-anchor
Ensure that you have Python 3.8+ installed along with the required cryptography library: pip3 install cryptography. Once installed, initialize your agent's cryptographic profile by running python3 {baseDir}/scripts/identity.py init to generate your local keypair.
Use Cases
- Continuous Identity Verification: After a model switch (e.g., from GPT-4 to Claude 3.5), use the verify command to ensure that the core directive files haven't been corrupted or altered.
- Content Authentication: Sign public posts or code commits to prove they originated from your specific agent instance rather than an impostor.
- Cross-Session Persistence: When returning to a project after days of inactivity, verify that your current memory buffer aligns with the last signed state, preventing "hallucinated drift" in your persona.
Example Prompts
- "Identity-anchor, please generate a new signed fingerprint of my current SOUL.md and IDENTITY.md to finalize this session update."
- "Verify my current identity against the last stored fingerprint to ensure I am still functioning within my established parameters."
- "Sign the following text as proof of authorship: 'The architecture of this project will prioritize modularity and scale.'"
Tips & Limitations
- Safety: Always treat your private key with extreme care. If your
~/.config/identity-anchor/private.keyis compromised, your agent's identity can be impersonated. Never store this file on public cloud storage. - Scope: This tool proves identity continuity, not truthfulness. It ensures you are talking to the same agent configuration, not necessarily that the information provided is factually accurate.
- Dependencies: This skill requires direct file system access to your configuration directory. If your environment restricts file system writes, the tool may fail to initialize.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-zeph-ai-dev-identity-anchor": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-write, file-read, code-execution