model-supply-chain-security

Verify model artifact provenance and integrity across deployment pipelines.

4|Updated Apr 27, 2026
One-click install
npx skills add https://github.com/maruakshay/mii-ai-security --skill model-supply-chain-security-maruakshay
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: model-supply-chain-security
Source: https://github.com/maruakshay/mii-ai-security/tree/main/skills/model-supply-chain-security
Command: npx skills add https://github.com/maruakshay/mii-ai-security --skill model-supply-chain-security-maruakshay

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The skill helps security teams ensure the provenance and integrity of model artifacts across the entire deployment pipeline, from base models to adapters and tokenizers, preventing unverified or tampered components from entering production.

Core Features & Use Cases

  • Verify model provenance and cryptographic integrity for weights, adapters, tokenizers, and config files.
  • Enforce a gated release workflow with staged environments and an approved registry for all artifacts.
  • Perform end-to-end reviews of model supply chains, including documentation and vendor provenance.

Quick Start

Register and verify each model artifact in an approved registry, enforce checksums at load time, and gate promotions through staging and production.

Frequently Asked Questions about model-supply-chain-security

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify model provenance and artifact integrity across a deployment pipeline?

To verify model provenance and artifact integrity, you register model weights, adapters, and tokenizers in an approved registry and enforce cryptographic checksums at load time. This ensures only verified components are deployed across staging and production environments.

What is model supply chain security and when do I need signed manifests?

Model supply chain security prevents unverified or tampered components from entering production. You need signed manifests when deploying base models, adapters, and tokenizers to cryptographically verify artifact origins and enforce gated release workflows.

How do I enforce a gated release workflow for staging and production model deployments?

You enforce a gated release workflow by registering all model artifacts in an approved registry and requiring checksum verification before promotion. This guarantees only cryptographically verified weights, adapters, and tokenizers move from staging to production environments.

Does this model supply chain security approach work with adapters and tokenizers?

Yes, this model supply chain security approach verifies cryptographic integrity for weights, adapters, tokenizers, and config files. It performs end-to-end reviews of all deployment components to ensure vendor provenance and prevent tampered artifacts from entering production.

What is the best way to prevent unverified model weights from entering production?

The best way to prevent unverified model weights from entering production is enforcing checksums at load time and using signed manifests. Gated promotion ensures only artifacts verified through an approved registry are deployed to production environments.