What is Civitas?

Civitas is Erethiel's core technology - an OpenAI-compatible Proof of Civitas SDK that enables developers to build autonomous, fully on-chain verifiable AI agents with cryptographic trust embedded into every action.

How It Works

Civitas leverages a Trusted Execution Environment (TEE) architecture to securely execute LLM API calls, ensuring verifiability through cryptographic attestations. Each attestation is posted on-chain to Solana for transparency and integrity.

Cryptographic Proof Components

Every action performed by a Civitas-enabled agent generates a cryptographic proof consisting of:

1. Hash: Representing the input and output

2. Signature: Ensuring the integrity of the action

3. Attestation (Optional): Providing additional validation for security

Example Workflow

1. Request: The agent receives a prompt

"What is Civitas by Erethiel?"

2. Processing: The agent generates a secure response

"Civitas is an infrastructure for autonomous AI agents with verifiable actions."

3. Proof Generation: The response is accompanied by a cryptographic proof

{
  "hash": "d87ac...",
  "signature": "ab43d...",
  "attestation": "secure"
}

4. Verification: The proof is validated to confirm the response originated from the agent operating in a secure, tamper-proof environment.