Prompt Injection Defense · seb1n/awesome-ai-agent-skills
Harden AI agents against prompt injection attacks
Threat-models an AI agent, RAG pipeline, or tool-using workflow for prompt injection risks, then proposes defenses such as isolating untrusted content, constraining tool access, and protecting secrets. Useful when reviewing agent security or investigating a suspected injection incident.
Good for
- Review an AI agent's architecture for injection risks
- Isolate untrusted content from agent instructions
- Add prompt-injection regression tests for an agent
- Source repository
- seb1n/awesome-ai-agent-skills
- Category
- Coding
Open-source skills are maintained by their authors and listed as published, with attribution. Results depend on how well the skill fits your task and material.
A good place to start
Review our customer-support agent's architecture and help me harden it against indirect prompt injection from web pages and documents.
Make your next great thing.
Bring a question, a file, or an idea that’s not quite there yet.