cerone 1.1.0

Zero Trust Security for AI Agents

Cerone — Governance and Zero Trust Runtime for AI Agents

Powered by AZTP (Agent Zero Trust Platform)

Cerone gives every AI agent a cryptographic identity, validates that its actions align with its declared purpose, and produces an auditable runtime trail across identity, validation, governance, and delegated token exchange.

Most teams deploying agents in production still have weak runtime control over what those agents actually do. Cerone is built to fix that.

---

Install

The current SDK/package name in this repo is aztp.

pip install git+https://github.com/AnantDhavale/AZTP.git

If you are working locally:

git clone https://github.com/AnantDhavale/AZTP.git
cd AZTP
pip install -e .

---

Get Your Free API Key

Self-serve signup. No waitlist and no approval step.

curl -X POST https://aztp-homer-semantics.onrender.com/signup \
  -H "Content-Type: application/json" \
  -d '{"email": "[e-mail]", "password": "your-secure-password", "name": "Your Name"}'

You receive an sk_free_... key instantly.

Free tier currently includes:

• 5,000 validations per 30-day window
• free for the first 30 days from signup
• bring your own OpenAI / Anthropic / other model-provider key
• Cerone does not proxy or charge for model inference

Hosted signup and support:

• homersemantics.com
• [e-mail]

Hosted service terms:

• HOSTED_TERMS.md

---

Quick Start

import asyncio
from aztp.sdk import AZTPClient, AZTPAgent

async def main():
    client = AZTPClient(
        api_url="https://aztp-homer-semantics.onrender.com",
        api_key="sk_free_...",
    )

    agent = AZTPAgent(
        purpose="Customer billing support",
        capabilities=["db_read", "billing_api"],
        client=client,
    )
    await agent.initialize()

    print(f"Agent ID: {agent.agent_id}")
    print(f"Trust score: {agent.certificate.trust_score}")

    result = await agent.validate_and_execute(
        tool="database_query",
        parameters={"table": "billing", "customer_id": "123"},
        execute_fn=lambda: {"amount": 99.99, "status": "paid"},
    )
    print(f"Result: {result}")

    child = await agent.spawn(
        purpose="Verify roaming charges",
        capabilities=["network_api"],
    )

    await agent.terminate()
    await client.close()

asyncio.run(main())

---

What Cerone Validates

Check	What it catches
Cryptographic identity	Impersonation, spoofed agents
Semantic alignment	Agents acting outside their declared purpose
Trust scoring	Behavioural drift over time
Capability scope	Agents calling tools they were never granted
Lineage integrity	Unauthorized parent-child relationships

---

Bring Your Own Model Key

Cerone governs agent behaviour, not inference.

You keep your own OpenAI, Anthropic, or other provider key and pass it directly to your model calls. Cerone validates the agent action and records the governance trail, but it does not sit in the middle of your model billing path.

import openai
from aztp.sdk import AZTPClient, AZTPAgent

client = AZTPClient(
    api_url="https://aztp-homer-semantics.onrender.com",
    api_key="sk_free_...",
)
openai_client = openai.AsyncOpenAI(api_key="sk-...")  # your key, your spend

agent = AZTPAgent(
    purpose="Summarise support tickets",
    capabilities=["read_ticket", "write_summary"],
    client=client,
)
await agent.initialize()

await agent.validate_and_execute(
    tool="write_summary",
    parameters={"ticket_id": "T-001"},
    execute_fn=lambda: openai_client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": "Summarise ticket T-001"}],
    ),
)

---

Decorator Pattern

agent = AZTPAgent(
    purpose="Customer data analysis",
    capabilities=["db_read", "analytics"],
    client=client,
)
await agent.initialize()

@agent.action("database_query")
async def query_customers(customer_id: str):
    return {"customer_id": customer_id, "name": "Jane Doe"}

customer = await query_customers("123")

---

Tiers

	Free	Startup	Pro	Enterprise
Validations / 30-day window	5,000	50,000	500,000	Custom
Agents	5	25	250	Custom
Audit retention	7 days	30 days	90 days	365 days
Model access	BYO only	BYO only	BYO now, managed later	BYO or managed
Support	Community	Email	Priority	Dedicated
Price	Free for first 30 days	Contact / self-serve pricing	Contact / self-serve pricing	Contact us

If you want fixed public pricing in this README, update this table once the commercial page is final.

---

Architecture

Your Agent Code
      │
      ▼
  Cerone SDK  ──────────────────────────────────────────┐
      │                                                  │
      ▼                                                  ▼
AZTP Platform (aztp-homer-semantics.onrender.com)  Your LLM Provider
  ├─ Identity Manager
  ├─ Semantic Validator
  ├─ Trust Engine
  └─ Audit Logger

---

Supported Frameworks and Integrations

Cerone currently ships adapters or normalization paths for:

• CrewAI
• Google ADK
• Gemma
• Salesforce
• ServiceNow
• Slack
• Microsoft 365
• Google Workspace
• Jira

---

Usage and Quota

curl https://aztp-homer-semantics.onrender.com/usage \
  -H "X-API-Key: sk_free_..."

This returns current usage, remaining quota, reset date, free-tier expiry, and tier feature flags.

---

Documentation

• SDK Onboarding Guide
• Hosted Service Terms
• Render Environment Changes

Live API docs:

• aztp-homer-semantics.onrender.com/docs

---

License

Current repository/package metadata is MIT.

The open-source repository license and the hosted Cerone service terms are separate:

• repository/package code: MIT
• hosted service usage: HOSTED_TERMS.md Free trial is subject to change. Use the software at your own risk.

---

Contact and Feedback

• Website: homersemantics.com
• API docs: aztp-homer-semantics.onrender.com/docs
• Support: [e-mail]
• Founder: [e-mail]
• Issues: github.com/AnantDhavale/AZTP/issues

If you are using Cerone, feedback is genuinely useful. I am doing some additions/ changes, please do reach out if you face any issues.