OraclesTech/guardian-sdk — Ethicore Engine™ is an AI safety, ethics, and compliance platform.

Description and use

**Brief Summary (2‑3 sentences)**
The Guardian SDK is the open‑source core of Ethicore Engine™, a multi‑layer safety platform that shields AI applications from prompt injection, jailbreaks, role hijacking, system‑prompt extraction and more than 100 other threat vectors. Built in Python, the SDK offers ready‑to‑use APIs, a CLI, and language‑agnostic metadata to integrate threat‑analysis directly into your models, RAG pipelines, or autonomous agents. With active maintenance, recent commits, and a growing community (≈ 105 ★), it’s a practical way to add robust AI security without building a safety stack from scratch.

**Value**

• **Turnkey safety**: Provides out‑of‑the‑box detection for a wide range of adversarial attacks, letting developers focus on product features rather than security research.
• **Accelerated AI development**: By exposing simple SDK/CLI hooks, teams can prototype new AI capabilities, RAG workflows, or agentic systems while keeping compliance and ethics checks baked in.
• **Cost‑effective compliance**: Reduces the need for custom rule engines or third‑party audits, helping meet internal governance and regulatory requirements faster.

**Practical Adoption Path**

1. **Evaluation** – Clone the repo, run the provided CLI against a sample prompt set to see detection rates and understand the threat categories covered.
2. **Integration** – Add the Python package to your project (pip install guardian-sdk), wrap your model inference calls with the SDK’s analyze() function, and configure any domain‑specific policies via the supplied JSON/YAML policy files.
3. **Testing & Tuning** – Use the built‑in logging and signal outputs to fine‑tune false‑positive thresholds, and optionally extend the policy set for proprietary threat models.
4. **Deployment** – Deploy the SDK as a sidecar service or embed it in your inference container; the lightweight design makes it suitable for both cloud functions and on‑premise environments.

**Production Readiness**

• **Active maintenance**: Last commit on 2026‑06‑27, regular issue triage, and a modest but engaged contributor base.
• **Maturity signals**: 105 GitHub stars, 11 forks, and a clear API/CLI surface indicate community validation.
• **Security posture**: No known critical vulnerabilities; the code focuses on detection rather than executing untrusted code, lowering attack surface.
• **Ecosystem fit**: Pure‑Python implementation integrates smoothly with popular LLM frameworks (LangChain, Haystack, OpenAI SDK) and can be wrapped for other languages via gRPC/REST.

Overall, Guardian SDK is production‑ready for pilots and early‑stage deployments, offering a high‑impact security layer with a low integration overhead.

Problem

OraclesTech/guardian-sdk helps add AI capability without starting from a blank model stack.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• prototype AI features
• build RAG or agent workflows
• evaluate model tooling

Evidence

Ethicore Engine™ is an AI safety, ethics, and compliance platform. This repo consists of the open-source components of Ethicore Engine™ - Guardian SDK; designed to protect your AI applications from prompt injection, jailbreaks, role hijacking, system-prompt extraction, and 100+ additional threat categories through a multi-layer analysis pipeline

Source