Tools
VectorGuard Nano
By VectorGuard plugin for OpenClaw
Configuration Example
import { secureSend, secureReceive } from './Vgn.js';
// Agent A sends
const payload = secureSend("Deploy model v2", "shared-secret", "agent-ops");
// → { encoded: "㤴㜷㐸㤷…", timestamp: 1738123456, note: "..." }
// Agent B receives
const original = secureReceive(
payload.encoded,
"shared-secret",
"agent-ops",
payload.timestamp
);
// → "Deploy model v2"README
# VectorGuard-Nano
**Messaging Security for AI Agents**
[](LICENSE)
[](https://openclaw.ai)
Simple, reversible HMAC-based string obfuscation for AI agents. Useful for basic message privacy in chat platforms, public channels, and prototyping.
> **Note**: This is VectorGuard-Nano—a lightweight implementation for simple use cases.
> For production-grade security with model-derived cryptography, see **[VectorGuard Enterprise](https://www.active-iq.com/vectorguard)**.
---
## What It Does
VectorGuard-Nano provides **deterministic string obfuscation** using:
- HMAC-SHA256 for generating shift patterns
- Shared secrets between agents
- Timestamp/nonce binding
- Reversible character shifting
**What it's NOT:**
- ❌ Not end-to-end encryption (use TLS for that)
- ❌ Not model-bound cryptography (see VectorGuard Enterprise)
- ❌ Not suitable for highly sensitive production data
- ❌ Not a replacement for proper auth/identity systems
**What it IS:**
- ✅ Quick obfuscation for agent-to-agent chat
- ✅ Traffic pattern protection for casual messaging
- ✅ Zero external dependencies (~100 lines)
- ✅ Drop-in for OpenClaw, Moltbook, Slack bots
---
## Quick Start
```js
import { secureSend, secureReceive } from './Vgn.js';
// Agent A sends
const payload = secureSend("Deploy model v2", "shared-secret", "agent-ops");
// → { encoded: "㤴㜷㐸㤷…", timestamp: 1738123456, note: "..." }
// Agent B receives
const original = secureReceive(
payload.encoded,
"shared-secret",
"agent-ops",
payload.timestamp
);
// → "Deploy model v2"
```
---
## Installation
### For OpenClaw Agents
```bash
# Clone repo
git clone https://github.com/Active-IQ/VectorGuard-Nano.git
# Copy to skills directory
cp -r VectorGuard-Nano ~/.openclaw/skills/vectorguard-nano
# Test
openclaw skill test vectorguard-nano
```
### Standalone (Node.js / Deno / Bun)
```bash
# Just copy Vgn.js to your project
curl -O https://raw.githubusercontent.com/Active-IQ/VectorGuard-Nano/main/Vgn.js
```
---
## Limitations
### Message Size
- **Recommended**: Under 1KB per message
- **Maximum**: ~4KB (technical limit from stream buffer)
- For larger data, split into chunks or use VectorGuard Enterprise
### Security Model
- Uses shared secrets (pre-shared key requirement)
- Timestamp-based nonce (5-minute drift tolerance by default)
- Deterministic (same input + timestamp = same output)
- No forward secrecy
- Obfuscation, not encryption
**For threat models requiring:**
- Model-derived keys
- Forward secrecy
- Key rotation
- Enterprise audit trails
- Compliance (SOC2, HIPAA, etc.)
**→ Upgrade to [VectorGuard Enterprise](https://www.active-iq.com/vectorguard)**
---
## Use Cases
### ✅ Good For:
- Agent chat on Moltbook, Discord, Telegram
- Obfuscating commands in public logs
- Prototype/demo applications
- Learning agent security concepts
- Non-sensitive inter-agent messaging
### ❌ Not For:
- Financial transactions
- Healthcare data (HIPAA)
- Personal identifiable information (PII) at scale
- Mission-critical production systems
- Replacing TLS/HTTPS
---
## Licensing
**Apache 2.0** for:
- Personal projects
- Open-source projects
- Educational/research use
**Commercial use** (production deployments, commercial products, revenue-generating services):
- Contact: **[email protected]**
See [LICENSE](LICENSE) for full terms.
---
## VectorGuard-Nano vs. Enterprise
| Feature | Nano (This Repo) | Enterprise |
|---------|------------------|------------|
| **Price** | Free (non-commercial) | Contact Sales |
| **License** | Apache 2.0 | Proprietary |
| **Crypto** | HMAC-SHA256 obfuscation | Model-derived keys |
| **Message Size** | < 1KB recommended | Unlimited |
| **Forward Secrecy** | No | Yes |
| **Key Management** | Shared secrets (manual) | Automatic + rotation |
| **Audit Trail** | Basic | Enterprise-grade |
| **Support** | Community (GitHub) | SLA-backed |
| **Compliance** | None | SOC2, HIPAA ready |
**Upgrade:** https://www.active-iq.com/vectorguard
---
## Contributing
We welcome contributions! Please:
1. Fork the repo
2. Create a feature branch
3. Submit a PR with clear description
For security issues, email: [email protected]
---
## About Active-IQ
Active-IQ builds AI security infrastructure. VectorGuard-Nano is our open-source contribution to the agent ecosystem.
**Learn more:**
- Website: https://www.active-iq.com
- Blog: https://www.active-iq.com/blog
- Enterprise: https://www.active-iq.com/vectorguard
---
**Built with ❤️ for the AI agent community**
tools
Comments
Sign in to leave a comment