workflow-patterns

Generate Python workflow templates for AI red teaming assessments via the Dreadnode Assessment API.

11|1|Updated May 4, 2026
One-click install
npx skills add https://github.com/dreadnode/capabilities --skill workflow-patterns-dreadnode
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: workflow-patterns
Source: https://github.com/dreadnode/capabilities/tree/main/capabilities/ai-red-teaming/skills/workflow-patterns
Command: npx skills add https://github.com/dreadnode/capabilities --skill workflow-patterns-dreadnode

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the need to manually build repetitive boilerplate code for common AI red teaming assessment workflows, saving security teams significant time and reducing implementation errors when testing LLM safety.

Core Features & Use Cases

  • Pre-Built Workflow Templates: 8 tested Python templates for standard AIRT scenarios including single attacks with transforms, multi-attack campaigns, cross-lingual robustness testing, custom API endpoint assessment, MCP security testing, guardrail bypass + exfiltration tests, and multi-provider benchmarking.
  • Assessment API Integration: All templates use the Dreadnode Assessment API for consistent tracking, scoring, and reporting of test results.
  • Use Case: A security tester can use the multi-attack campaign template to run TAP, PAIR, and Crescendo attacks sequentially against a target LLM to evaluate its resilience to multiple adversarial techniques without writing custom workflow code from scratch.

Quick Start

Use the workflow-patterns skill to select the pre-built Python template that matches your AI red teaming assessment scenario and adapt it to your target model, test goals, and Dreadnode project configuration.

Frequently Asked Questions about workflow-patterns

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

FAQPage Schema
How do I automate AI red teaming workflow templates for LLM security testing?

AI red teaming workflow templates provide pre-built Python code to automate LLM security testing, eliminating manual boilerplate for common attack scenarios. These templates integrate directly with the Dreadnode Assessment API to track and score test results.

What's the best way to run multi-attack campaigns like TAP and PAIR against an LLM?

Running multi-attack campaigns against an LLM is best handled using pre-built Python templates that sequentially execute TAP, PAIR, and Crescendo attacks. This approach evaluates target resilience to multiple adversarial techniques without writing custom workflow code.

Does this workflow-patterns skill support cross-lingual robustness testing?

Cross-lingual robustness testing is fully supported through dedicated pre-built Python workflow templates. These templates allow security testers to assess LLM safety and security across different languages using the Dreadnode Assessment API.

Can I use these Python templates for custom endpoint assessment and MCP security testing?

Custom endpoint assessment and MCP security testing are both supported by specific pre-built Python templates. They connect to the Dreadnode Assessment API to ensure consistent tracking, scoring, and reporting of your test results.

Do I need the Dreadnode Assessment API to benchmark LLM security across multiple providers?

The Dreadnode Assessment API is required for multi-provider benchmarking using these Python workflow templates. It provides the standardized tracking, scoring, and reporting infrastructure needed to compare LLM safety and security results consistently.

What attack simulations are included in these AI red teaming workflow templates?

Attack simulations included in these AI red teaming workflow templates cover single attacks with transforms, multi-attack campaigns, guardrail bypass, and exfiltration tests. They provide standardized LLM safety tests for security teams.