pentest-ai-redteam
CommunityAI red-team to strengthen secure agents
Software Engineering#pentest#ai-security#prompt-injection#redteam#tool-call-abuse#adversarial-examples
AuthorXerrion
Version1.0.0
Installs0
System Documentation
What problem does it solve?
AI/LLM red-team testing frames and standardizes evaluation of prompt-injection vulnerabilities, tool-call abuse, and resilience against adversarial techniques.
Core Features & Use Cases
- Comprehensive taxonomy covering direct and indirect prompt injections, jailbreaking, role confusion, instruction override, system-prompt extraction, and defense-aligned probe patterns.
- Attack library and probes providing reproducible test scenarios, evidence workflows, and mappings to OWASP LLM Top 10 and MITRE ATLAS for AI-security assessments.
- Structured evaluation with deterministic transcripts, canaries, and secure logging to the .pentest/evidence directory.
Quick Start
Kick off a red-team engagement by selecting an AI system and following the playbook to enumerate injection vectors and produce reproducible evidence.
Dependency Matrix
Required Modules
None requiredComponents
Standard package💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: pentest-ai-redteam Download link: https://github.com/Xerrion/opencode/archive/main.zip#pentest-ai-redteam Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.