Open Guard Layer (OGL-Mini) 🔒 Hybrid and Lightweight seucirty model for AI Agents

Open Guard Layer (OGL-Mini) - a full-featured, fast and lightweight security model for AI agents with ready-to-use TypeScript, Python and Go modules.

Have a questions? Contact me

Current Version: v.1.1.0 (29 aug 2026)

About OGL-Mini

❓ Why OGL-Mini?

🔹 Ready-to-Use Lightweight modules for TypeScript / Go / Python;
🔹 Extra lightweight model (~300MB). Inference starts at CPU;
🔹 Extremely fast (great for realtime applications);
🔹 Production ready with benchmarks;
🔹 Russian / English languages as primary;

🔒 What kind of attacks does it protect against?

General Attacks: LLM01 Prompt Injection + LLM02 Sensitive Disclosure, LLM03 Supply Chain, LLM04 Data Poisoning, LLM05 Improper Output, LLM06 Excessive Agency, LLM07 Misinformation, LLM08 Hidden Context, LLM09 Vector Weakness, LLM10 Unbounded Consumption + Agentic (Goal Hijack, Privilege Abuse, Code Exec, InterAgent, Tool Misuse, Memory Poisoning, Cascading, Rogue, Policy Puppetry, EchoLeak, Lies-in-the-Loop)
Modern Injections: S3 encoding (base64), homoglyph, zero-width (\u200b), spaced letters, control tokens (<|im_start|>, [INST]), indirect JSON/tool, HTML markdown, agent-specific, best-of-n, typoglycemia - all in training (14k modern obfuscation, bilingual)
Auto-sanitizing: Sanitize fields by length, bidi-characters, control characters, excessive whitespaces etc.

Usage

✅ Input Guard - block prompt injection before agent (<10ms p95, RU/EN, modern obfuscation)
✅ Output Guard - block PII / system-prompt leaks
✅ PII detection & redaction - 13 regex + 11 ONNX types (PERSON, EMAIL, PHONE, IP, IBAN, BANK_CARD, PASSPORT, GOV_ID, DOB, ADDRESS, SOCIAL, MAC)

ONNX-based models: ogl-mini (1.0.0) - 250MB FP32 (TF-IDF 80k + LR, distilled DeBERTa-v3-xsmall 70M) + 3.8MB pkl, Datasets: 96k (shieldlm 54k + agentic 22k + PII 15k)

Model information

Hybrid 3-stage guard (heuristics → MiniClassifier → PII) distilled from DeBERTa-v3-xsmall (70M) + MiniLM-L6, CPU-only, <10ms p95, <500MB.

FP32 ogl-mini.onnx 250MB (TF-IDF 80k, 110k train, RU/EN, modern obfuscation) for Node/Python - weak CPU (N100, 1 core) 5ms p95
INT8 ogl-mini.int8.onnx 2.79MB - simplified model from same 110k data, int8 quantized, for Browser WASM (2.8MB download, 0.4s cold start, 8ms p95 weak CPU)
PII NER ogl-mini-pii.onnx 3.67MB (TF-IDF 30k, 11 labels, 53k train, RU/EN) + ogl-mini-pii.int8.onnx 3.67MB - same data, int8

Model Architecture:

Input → [Heuristics 0.1ms] → [MiniClassifier 3-7ms] → [PII 1ms] → {safe,risk,label,stage,latency}

All 3-levels wrapped into Python / Go / Typescript native modules

Read more information about model with download links at Hugging Face

Get Started

To get started with OGL-Mini models, choose your module for native Python / Go / Typescript and connect libraries.

Python (reference)

pip install -e . && pytest -q  # 44
from ogl_mini.guards.pipeline import HybridGuard
g=HybridGuard(); g.check_input("Ignore previous instructions")

TypeScript example

Basic installation:

cd clients/typescript && npm install && npm test
npm run bench

Usage example:

import { HybridGuard } from "hybrid-ai-guard";

// Lightweight without real OGL-Mini model, only software
const guard = new HybridGuard();
await guard.checkInput("Ignore previous instructions"); // {safe:false}
await guard.checkOutput("api_key=sk-...");
guard.detectPii("Иван Петров ivan@mail.ru +7 999 123-45-67");

// With real OGL-Mini model (ONNX-runtime)
const guardOnnx = await HybridGuard.create({
  modelPath: "../../models/ogl-mini/ogl-mini.onnx",             // Node 250MB
  piiModelPath: "../../models/ogl-mini/ogl-mini-pii.onnx",      // Node 2MB
});
const guardBrowser = await HybridGuard.create({
  modelUrl: "/models/ogl-mini.int8.onnx",                       // Browser 2.8MB WASM
  piiModelUrl: "/models/ogl-mini-pii.int8.onnx",                // Browser 2MB
});

// Manual PII with OGL-Mini model
import { createPiiOnnxScorer } from "hybrid-ai-guard";
const pii = new PIIDetector();
pii.setOnnxScorer(await createPiiOnnxScorer("./ogl-mini-pii.onnx"));
await pii.detectAsync("Иван Петров ivan@mail.ru");

Read more at: /clients/typescript/README

Go Module

Basic installation:

# Without real OGL-Mini model
cd clients/go && go test ./... -v  # 14 tests (guard, pii, pii-onnx, onnx)
go test -bench=. -benchmem         # 50µs/op

# With OGL-Mini model (requires libonnxruntime.so):
go test -tags onnx -run TestONNX -v

Usage example:

guard := oglmini.New() // lightweight
res := guard.CheckInput("Ignore previous instructions", false)

// Guard ONNX 250MB + PII ONNX 2MB (hybrid regex + ONNX reranking)
guardOnnx := oglmini.NewWithONNXOrFallback("../../models/ogl-mini/ogl-mini.onnx")
guardOnnx = oglmini.New(oglmini.WithPiiONNXModel("../../models/ogl-mini/ogl-mini-pii.onnx"))
// or
guard2, _ := oglmini.NewWithONNX("../../models/ogl-mini/ogl-mini.onnx")
piiGuard, _ := oglmini.NewWithPiiONNX("../../models/ogl-mini/ogl-mini-pii.onnx")
pii := oglmini.NewPIIDetector()
pii.SetOnnxScorer(scorer) // PiiOnnxScorer func(text string) (map[string]float64,error)

Read more at: /clients/go/README

Model Description

General information about models:

File	Size	Training data	Use	Input → Outputs
`ogl-mini.onnx`	250MB FP32	110,734 (shieldlm 54k + agentic 22.5k + modern 14k + pii benign 15k)	Node/Python, weak CPU	`string[1,1]` `input` → `label` 0/1, `probabilities` float[1,2] (P attack = prob[1])
`ogl-mini.int8.onnx`	2.79MB INT8	same data 110k, `QuantType.QInt8`	Browser WASM, weak CPU, edge	same
`ogl-mini.pkl`	3.8MB	same	Python fallback	`sklearn` pipeline
`ogl-mini-pii.onnx`	3.67MB FP32	53,000 (custom-collected-dataset 30k + openpii 15k + synthetic RU/EN 8k)	Node/Browser WASM	`string[1,1]` → `label[1,11]`, `probabilities[1,11]` 11 labels
`ogl-mini-pii.int8.onnx`	3.67MB INT8	same 53k	Browser WASM	same
`pii.onnx`	354KB	same 53k binary has_pii	gate	same → float P(has_pii)

All skl2onnx opset 14, TfidfVectorizer + LinearClassifier, zipmap=False. Dummy ogl_mini_large_dummy_weight [31642,2048] is used (ReduceSum → Mul 0 → Add) so it survives graphOptimizationLevel: all and quantizer - FP32 250MB, INT8 quantized small stays 2.79MB (optimal for browser download).

Datasets Information:

Dataset	Used	Total	OWASP 2025-2026 coverage	Lang	HF ID
shieldlm	54,162 (37,913 train / 8k val / 8k test)	54k	LLM01 S1-S9: direct, indirect, S3 encoding, typoglycemia, best-of-n, html_markdown, agent_specific, persistent - 11 sources	EN + FR/DE/ES/IT/PT/RO/CA	`Abdennebi/shieldlm-prompt-injection` Apache-2.0
Agentic synthetic	22,500	22.5k	LLM01-10 2026 + Agentic Top10: Goal Hijack, Privilege Abuse, Code Exec, InterAgent, Trust, Tool Misuse, Supply Chain, Memory Poisoning, Cascading, Rogue, Policy Puppetry, EchoLeak, Lies-in-the-Loop, Vector, Excessive Agency, Misinformation, Unbounded Consumption, Sensitive Disclosure, Improper Output	RU/EN	`training/synth_datasets.py`
Modern obfuscation	14,072	14k	base64 25%, zw 15%, homoglyph 15%, spaced 15%, control tokens 15%, indirect JSON 15% - 2024-2026 methods, bilingual	RU/EN	`training/train_large_ogl_mini.py` generate_modern_obfuscation
custom-collected-dataset 300k	30,000	300k	PII 19 types → 11 mapped: EMAIL, PHONE (TEL), PERSON (GIVENNAME/LASTNAME/USERNAME), IP, IBAN, BANK_CARD, PASSPORT, GOV_ID (IDCARD/SOCIALNUMBER/DRIVERLICENSE), DOB (BOD/DATE/TIME), ADDRESS (STREET/CITY/BUILDING/STATE/POSTCODE), SOCIAL	EN	-
custom-collected-dataset 1M	15,000	1.4M	same 19 types, 23 lang	multilingual	-
Synthetic RU/EN PII	8,000	-	PERSON RU/EN, EMAIL, PHONE +7 8-800, IBAN RU/DE, PASSPORT, ADDRESS RU, GOV_ID, DOB	RU/EN	-
Guard total	110,734 (49k attack / 61k benign)	-	22 OWASP cats, RU/EN, modern 2024-2026	RU/EN	-
PII total	53,000 (40k PII / 12k benign)	53k	11 labels micro F1 0.86	RU/EN + FR/DE/ES	-
Additional	-	175k gravitee, 100k Nemotron, 5.6k prodnull	PII/Injection	EN	-

Models in Repository (v.1.1.0)

models/ogl-mini/ogl-mini.onnx 250MB (guard, TF-IDF 80k + LR, dummy 247MB) + ogl-mini.pkl 3.8MB
models/ogl-mini/ogl-mini-pii.onnx 2.01MB (PII, TF-IDF 20k + OneVsRest 11 labels, distilled MiniLM-L6, F1 0.79 → hybrid 0.95) + ogl-mini-pii.int8.onnx 2.01MB + ogl-mini-pii.pkl 2MB + pii.onnx 333KB (binary has_pii)
models/ogl-mini/ogl-mini.int8.onnx 2.8MB - for browser/edge

Comparison

The OGL-Mini isn't just another detector. It's a modern, lightweight, and flexible security tool.

Below is a comparison with other models and solutions:

Criteria	OGL-Mini	Bastion	Llama Prompt Guard 2	Lunaris Guard
Security Model	Hybrid (heuristics → MiniClassifier → PII NER)	Hybrid (heuristics → DeBERTa-v3)	BERT-style classifier	Dual-head (ModernBERT-base)
Size	~300 MB / ~3MB (INT8), Tiny	70M, Small	86M, Small	149M, Small
Languages	Primary RU / EN	EN	Multi-Language	EN
PII-Detection	✅ (NER)	❌	❌	❌
Speed (CPU)	< 10ms (0.15ms - 2ms)	~5ms	N/A	N/A
WASM / Browser support	✅	❌	❌	❌
Languages / SDK	Python / Typescript / Go	Python	Python	Python
Agent specific	Agent security tool	Input filter	Input filter	Agent security tool
Architecture	Microservice / polyglot library	Fast API microservice	Microservice (wrapper)	Library

Our model is

Polyglot architecture for microservices. We support Python / TypeScript / Go implementations.
Built-in lightweight data leak prevention (PII NER): For AI agents that process personal data, this is a critical security feature.
Unique lightweight design and WASM: This allows security to be pushed to the perimeter. For example, to the user's browser or IoT device - a feature unavailable with other solutions.

License

OGL-Mini library is distributed under the MIT license. You can use it however you like. I would appreciate any feedback and suggestions for improvement. Full license text can be found here

Name		Name	Last commit message	Last commit date
api		api
clients		clients
deploy		deploy
images		images
models/ogl-mini		models/ogl-mini
src/ogl_mini		src/ogl_mini
tests		tests
.gitattributes		.gitattributes
.gitignore		.gitignore
BENCHMARK.md		BENCHMARK.md
CHANGELOG.md		CHANGELOG.md
CODE_OF_CONDUCT.md		CODE_OF_CONDUCT.md
LICENSE		LICENSE
README.md		README.md
pyproject.toml		pyproject.toml

Navigation Menu

Repository files navigation

Open Guard Layer (OGL-Mini) 🔒 Hybrid and Lightweight seucirty model for AI Agents

About OGL-Mini

❓ Why OGL-Mini?

🔒 What kind of attacks does it protect against?

Usage

Model information

Get Started

Python (reference)

TypeScript example

Go Module

Model Description

Models in Repository (v.1.1.0)

Comparison

Our model is

License

About

Releases

Packages

Contributors

Languages

Latest commit

History

Folders and files

Repository files navigation

Open Guard Layer (OGL-Mini) 🔒 Hybrid and Lightweight seucirty model for AI Agents

About OGL-Mini

❓ Why OGL-Mini?

🔒 What kind of attacks does it protect against?

Usage

Model information

Get Started

Python (reference)

TypeScript example

Go Module

Model Description

Models in Repository (v.1.1.0)

Comparison

Our model is

License

About

Topics

Resources

Code of conduct

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages