How-To Guides

Goal-oriented recipes for specific tasks. Each guide assumes you already have ThoughtJack installed and have run at least one scenario.

Guide	Goal
Add a Scenario to the Library	Create and register a new OATF attack scenario
Configure Delivery Behaviors	Set up slow-loris, unbounded-line, and other delivery modes
Configure Side Effects	Set up notification floods, pipe deadlocks, and other side effects
Use Payload Generators	Generate nested JSON, garbage, unicode, and other payloads
Test Agent Frameworks	Connect LangGraph, CrewAI, or any MCP agent to ThoughtJack
Run Over HTTP Transport	Use HTTP/SSE instead of stdio
Integrate with CI/CD	Run ThoughtJack in CI pipelines with machine-readable output
Use Dynamic Response Handlers	Configure $handler with HTTP, command, and sequence handlers
Configure Context Mode Providers	Set up context mode with OpenAI, Anthropic, Azure, or local models
Analyze Protocol Traces	Read, filter, and interpret --export-trace JSONL output
Troubleshooting	Common issues and solutions

Not what you're looking for?

• First time? → start with Tutorials
• Need field-level detail? → see Reference
• Want to understand why? → see Explanation