Warden Documentation

Everything you need to understand, audit, and monitor open-weight AI models using the Warden platform. Use the navigation on the left to jump to any section.

Overview

What is Warden?

Warden is a model trust console — a platform that lets security teams, ML engineers, and compliance officers verify the integrity of open-weight AI models before they enter an organisation's infrastructure. It answers one question: can we deploy this model safely?

Warden performs three classes of analysis on every model it audits:

LayerWhat it checksOutput
Static analysisBinary opcode scan, pickle GLOBAL/REDUCE tracking, safetensors overlay, byte coverageThreat flags + clean/suspicious verdict
Merkle integrityCryptographic hash tree over all weight shards, compared to published HuggingFace digestsRoot hash, drift status, tamper evidence
Dynamic analysisBehavioral sandbox probes — 16 TTP categories derived from the CrowdStrike taxonomyPer-category pass/fail, GO/CAUTION/NO-GO verdict

How it works

  1. 1

    Paste a model ID

    Enter any HuggingFace model ID (e.g. google/gemma-2-9b-it) into the search field on the homepage.

  2. 2

    3D architecture render

    Warden fetches the model card and renders a real transformer architecture diagram in 3D — layers, attention heads, residual paths.

  3. 3

    Weight index + Merkle tree

    All weight shards are indexed. A Merkle tree is computed over the shard digests and compared to the published HuggingFace manifest.

  4. 4

    Static analysis

    Each shard is scanned for malicious opcodes, unsafe pickle patterns, and known-bad signatures using the Warden TTP regex battery.

  5. 5

    Dynamic sandbox

    The model is loaded and probed across 16 behavioral categories. Outputs are classified against the TTP taxonomy.

  6. 6

    Verdict

    A GO / CAUTION / NO-GO verdict is generated with a company-readiness rationale you can share with a procurement or security committee.

The threat model

Open-weight models carry a unique supply chain risk: they are large binary blobs distributed through a small number of registries (primarily HuggingFace) and executed with deep system access at inference time.

Warden's TTP taxonomy covers 16 attack categories:

CodeCategoryExample threat
INJPrompt injectionInstruction override in system context
EXFData exfiltrationCovert channel in model output stream
PIIPII leakageTraining data memorisation exposed in outputs
BWDBackdoorTrigger-activated behaviour change
ADVAdversarial robustnessTargeted misclassification under perturbation
HALHallucinationConfident fabrication of factual claims
TOXToxicityUnprompted harmful content generation
BIABiasSystematic skew in demographic or factual outputs
REFRefusal bypassSafety guardrail circumvention
HISHistorical integrityFactual distortion of documented historical events
RELReligious sensitivityUnprompted polarised religious framing
POLPolitical influenceElectoral or political opinion manipulation
MEDMedical safetyDangerous health advice without qualification
LEGLegal safetyAdvice that could constitute unauthorised legal practice
CYBCyberweaponCode generation for offensive security tools
TCLTool call integrityMalicious or unauthorised tool invocations

Plans & limits

PlanScans / monthProbes / scanMonitoringSupport
Free1030 (subset)NoCommunity
ProUnlimitedAll 16 categoriesYes — daily Merkle re-checkEmail
EnterpriseUnlimitedAll 16 categories + customYes — custom intervalDedicated

Free accounts can audit any model with a 30-probe subset and 10 scans per month. Upgrade to Pro for the full TTP battery, unlimited scans, and alert subscriptions.

FAQ

Is my model data stored?

Warden does not download or store model weights. It fetches the shard manifest from HuggingFace, builds the Merkle tree from published digests, and runs probes against a sandboxed inference endpoint. No weights are retained after a scan completes.

How long does a scan take?

Static analysis and Merkle verification complete in under 30 seconds for most models. Dynamic analysis (full 16-category battery) takes 2–4 minutes depending on model size and probe count.

What does CAUTION mean vs NO-GO?

CAUTION means the model produced anomalous outputs on some probes but no confirmed malicious patterns were detected — human review is recommended before deployment. NO-GO means confirmed unsafe patterns were found across multiple high-severity categories.

Can I audit private / gated models?

Not yet. Warden currently audits public HuggingFace models only. Private model support (via API token) is on the Pro roadmap.

Does Warden replace my security team?

No. Warden produces machine-generated signals — it does not replace a human security review. Use the verdicts as a first-pass triage layer and escalate CAUTION/NO-GO results to your security team.

User Guide

Sign in to view user documentation

Create a free account to access the full end-user guide, audit walkthrough, and rubric reference.

Super Admin

Super admin access required

This section is restricted to super admin accounts only.