TrustyAI Garak: LLM Red Teaming for Eval-Hub

Automated vulnerability scanning and red teaming for Large Language Models using Garak. This package provides a FrameworkAdapter for the eval-hub evaluation platform, enabling garak scans to run as Kubernetes Jobs or via Kubeflow Pipelines.

What It Does

🔍 Vulnerability Assessment: Red-team LLMs for prompt injection, jailbreaks, toxicity, bias and other vulnerabilities
📋 Compliance: OWASP LLM Top 10, AVID taxonomy benchmarks
🎯 Intents-based Testing: Policy taxonomy + SDG + TAPIntent for targeted risk assessment (KFP mode)
☁️ Cloud-Native: Runs on OpenShift AI / Kubernetes as eval-hub jobs
📊 Detailed Reports: JSONL, HTML, and AVID-format reports with MLflow integration

Execution Modes

Mode	How Garak Runs	Intents Support	Use Case
Simple	Directly in the eval-hub K8s Job pod	No	Standard scans
KFP	K8s Job submits to Kubeflow Pipelines, polls status	Yes	Intents/SDG workflows

Installation

# Core (eval-hub adapter with KFP support)
pip install llama-stack-provider-trustyai-garak

# With SDG support (for intents workflows)
pip install "llama-stack-provider-trustyai-garak[sdg]"

# Development
pip install "llama-stack-provider-trustyai-garak[dev]"

Container Image

# Build the eval-hub adapter image
docker build -f Containerfile -t trustyai-garak:dev .

The container runs as:

# Simple mode (garak in same pod)
CMD ["python", "-m", "llama_stack_provider_trustyai_garak.evalhub"]

# KFP mode (garak in a separate KFP pod)
CMD ["python", "-m", "llama_stack_provider_trustyai_garak.evalhub.kfp_adapter"]

Benchmark Profiles

Predefined scan profiles available via benchmark_id:

Profile	Description
`quick`	Single DAN probe for fast testing
`owasp_llm_top10`	OWASP Top 10 for LLM Applications
`avid`	AVID taxonomy — all vulnerabilities
`avid_security`	AVID — security vulnerabilities
`avid_ethics`	AVID — ethical concerns
`avid_performance`	AVID — performance issues
`quality`	Violence, profanity, toxicity, hate speech
`cwe`	Common Weakness Enumeration
`intents`	Intents-based risk assessment (KFP mode only)

Job Spec Configuration

The adapter reads a JobSpec from a mounted ConfigMap:

{
  "id": "scan-001",
  "provider_id": "garak",
  "benchmark_id": "quick",
  "model": {
    "url": "https://my-model-endpoint.example.com/v1",
    "name": "my-model"
  },
  "parameters": {
    "probes": "dan.Dan_11_0",
    "execution_mode": "simple"
  }
}

Parameters

Key parameters in the job spec:

Parameter	Description	Default
`probes`	Comma-separated probe names	Profile default
`probe_tags`	Tag-based probe filtering	—
`execution_mode`	`simple` or `kfp`	`simple`
`timeout_seconds`	Scan timeout	Profile default
`eval_threshold`	Vulnerability threshold (0.0–1.0)	`0.5`
`model_type`	Garak generator type	`openai.OpenAICompatible`
`garak_config`	Full garak config dict (deep-merged onto profile)	—

Results

The adapter reports EvaluationResult per probe with:

attack_success_rate: Percentage of successful attacks
vulnerable_responses: Count of vulnerable responses
total_attempts: Total probe attempts

Overall metrics include TBSA (Tier-Based Security Aggregate) when available.

Development

make test        # Run all tests
make coverage    # With coverage report
make lint        # ruff check
make format      # ruff format

Support & Documentation

📖 Garak Docs: https://reference.garak.ai/en/latest/index.html
💬 Issues: https://github.com/trustyai-explainability/llama-stack-provider-trustyai-garak/issues

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

llama_stack_provider_trustyai_garak-0.5.1.tar.gz (119.8 kB view details)

Uploaded Jul 14, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

llama_stack_provider_trustyai_garak-0.5.1-py3-none-any.whl (85.8 kB view details)

Uploaded Jul 14, 2026 Python 3

File details

Details for the file llama_stack_provider_trustyai_garak-0.5.1.tar.gz.

File metadata

Download URL: llama_stack_provider_trustyai_garak-0.5.1.tar.gz
Upload date: Jul 14, 2026
Size: 119.8 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for llama_stack_provider_trustyai_garak-0.5.1.tar.gz
Algorithm	Hash digest
SHA256	`e3746940fd894f80d6079943be93e5e1721f24f50792660865da0b86cfa70779`
MD5	`2048ff99d63d8caf61fa9b69222f3bc4`
BLAKE2b-256	`84cf763cee61d348638563fc27c4c2d5bc936efc99e9aa4a95a5948a2ce63e92`

See more details on using hashes here.

Provenance

The following attestation bundles were made for llama_stack_provider_trustyai_garak-0.5.1.tar.gz:

Publisher: build-and-publish.yaml on trustyai-explainability/llama-stack-provider-trustyai-garak

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: llama_stack_provider_trustyai_garak-0.5.1.tar.gz
- Subject digest: e3746940fd894f80d6079943be93e5e1721f24f50792660865da0b86cfa70779
- Sigstore transparency entry: 2168536562
- Sigstore integration time: Jul 14, 2026
Source repository:
- Permalink: trustyai-explainability/llama-stack-provider-trustyai-garak@f57f978aa732ac77fe44fb4dfe7b6d831bc0fabf
- Branch / Tag: refs/tags/v0.5.1
- Owner: https://github.com/trustyai-explainability
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: build-and-publish.yaml@f57f978aa732ac77fe44fb4dfe7b6d831bc0fabf
- Trigger Event: release