CapKit - Capability Kit for AI Agents

Give your AI agents the minimum permission they need. No more, no less.

🚀 Quick Start (5 minutes)

git clone https://github.com/iamGodofall/capkit.git
cd capkit
npm install
npm run demo  # See it work!

Why CapKit?

Current AI agents have root access to your systems. One prompt injection = disaster.

CapKit issues cryptographically signed, time-bound capabilities:

• ✅ Scoped: post to /twitter, not delete
• ✅ Time-bound: Expires in 10min
• ✅ Signed: HMAC-SHA256 verification
• ✅ Auditable: Every action logged
• ✅ Zero deps: Sovereign-first (Node crypto only)

Security Model

Threat Model:
1. Agent prompt injection → Capability stays scoped
2. Key compromise → Time-bound limits damage  
3. Network failure → Works offline
4. Malicious actor → HMAC prevents tampering

Demo

See demo/agent-example.ts

Install

npm install capkit

API

import { issueCapability, verifyCapability, checkPolicy } from 'capkit';

// Issue scoped permission
const cap = issueCapability({
  action: 'post',
  resource: '/twitter',
  key: 'your-secret-key',
  expiresIn: '10m'
});

// Agent uses capability
if (verifyCapability({ capability: cap, key: 'your-secret-key' }).valid) {
  console.log('✅ Agent authorized');
}

---

Part of the Agent Builder Suite
→ capkit: Scoped capabilities for agents
→ quickbench: Reproducible agent evaluation
→ edge-run: Offline-first orchestration (coming soon)
→ connector-starter: Generate adapters fast (coming soon)

Built for builders who ship. MIT licensed. Local-first by design.

License

MIT

---

☕ Support the studio

Everything here is built indie — no VC money, no marketing budget. If this project saves you time or you believe in local-first, sovereign software, fuel the work:

Buy Me a Coffee · PayPal

Every contribution goes into development time and keeping these tools free & MIT-licensed.

— Enock Labs · enockgames.live