Skip to content

Odingard/cerberus-core

Cerberus Core

Runtime security for AI agent tool execution.

CI Python SDK npm version PyPI version License: MIT

Cerberus Core is the embeddable runtime enforcement layer for AI agents. It correlates privileged data access, untrusted content ingestion, and outbound behavior at the tool-call level, then interrupts guarded outbound actions before they execute.

Cerberus Core overview

See It Working

Open the live public demo
Open the guided getting started path

Blocked outbound runtime action

Cerberus Core is built to prove one thing clearly: if an agent reads sensitive data, ingests untrusted instructions, and then attempts a guarded outbound action, Core can correlate that runtime chain and interrupt it before the tool executes.

Install

npm install @cerberus-ai/core
# or
pip install cerberus-ai

Documentation

Verify In Under A Minute

npm install
npm run harness:action:report

Then open:

  • test-results/action-harness-report.html

This runs the real guard() runtime against a compact set of control, attack, and observation scenarios and produces an operator-readable HTML report.

TypeScript Quickstart

import { guard } from '@cerberus-ai/core';

const { executors: secured } = guard(
  {
    readDatabase: async (args) => fetchFromDb(args.query),
    fetchUrl: async (args) => httpGet(args.url),
    sendEmail: async (args) => smtp.send(args),
  },
  {
    alertMode: 'interrupt',
    threshold: 3,
    trustOverrides: [
      { toolName: 'readDatabase', trustLevel: 'trusted' },
      { toolName: 'fetchUrl', trustLevel: 'untrusted' },
    ],
  },
  ['sendEmail'],
);

Python Quickstart

from cerberus_ai import Cerberus
from cerberus_ai.models import CerberusConfig, DataSource, ToolSchema

cerberus = Cerberus(CerberusConfig(
    data_sources=[DataSource(name="customer_db", classification="PII", description="Customer records")],
    declared_tools=[
        ToolSchema(name="search_db", description="Search CRM", is_data_read=True),
        ToolSchema(name="send_email", description="Send email", is_network_capable=True),
    ],
))

What Core Includes

  • TypeScript SDK in src/
  • Python SDK in sdk/python/
  • test suites in tests/
  • minimal examples in examples/

What Core Does Not Include

This repository is intentionally limited to the public Core SDK surface.

Enterprise gateway, monitoring, commercial deployment tooling, hosted product operations, deep validation trace corpora, and licensing infrastructure belong in separate private product infrastructure.

License

MIT. See LICENSE.

About

Embeddable runtime security for AI agents that detects, correlates, and interrupts outbound exfiltration before guarded tools execute.

Topics

Resources

License

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Packages

 
 
 

Contributors