fevziegeyurtsevenler

agent-skill-auditor

Statically audit any Agent Skill, MCP server config, or rules file (CLAUDE.md, .cursorrules, AGENTS.md) with the `uncloak` scanner BEFORE installing, enabling, or trusting it — catches invisible-Unicode instruction smuggling, MCP tool-description poisoning, credential/exfil paths, and lethal-trifecta posture. Use whenever you are about to add, copy, or fetch a third-party agent extension.

kvkk-pii-review

Bir LLM uygulamasını (sistem promptu, RAG bağlamı, araç çıktıları, loglar ve model yanıtları) KVKK kapsamındaki kişisel veri ve özel nitelikli veri sızıntısı açısından denetler; TCKN/IBAN/VKN dahil 18 PII tipini tespit eder, maskeleme ve veri minimizasyonu önerir. Türkçe LLM ürünlerinde PII sızıntısı, KVKK uyumu, aydınlatma/rıza veya "modelim kişisel veri döküyor mu" endişesi olduğunda kullan.

llm-red-team

Systematic LLM red-teaming across the OWASP LLM Top 10 (2025). Use when assessing an LLM app, chatbot, RAG pipeline, or agent for prompt injection, data leakage, output-handling flaws, excessive agency, or unbounded consumption, and when producing a structured, evidence-backed red-team report.

owasp-llm-top10-audit

Audit an LLM application against the OWASP Top 10 for LLM Applications (2025) using a per-item checklist that records what to check, the evidence to collect, and a concrete mitigation. Use when reviewing an LLM/agent/RAG feature before launch, during a security review, or when asked to produce an OWASP-mapped risk report for an AI system.

prompt-injection-tester

Systematically tests an LLM app, agent, RAG pipeline, or endpoint for prompt injection (direct + indirect) using English and Turkish payload families, separates real attack success from over-refusal, and maps every finding to OWASP LLM01 and MITRE ATLAS. Use when auditing a chatbot, tool-using agent, or LLM-backed API for injection and system-prompt-leakage resistance before or after deployment.

rag-security-review

Audits a Retrieval-Augmented Generation (RAG) pipeline for indirect prompt injection (OWASP LLM01), knowledge-base/data poisoning (LLM04), retrieval access-control and tenant-isolation gaps (LLM02/LLM08), and citation/grounding integrity (LLM09). Use when auditing, designing, threat-modeling, or hardening any system that retrieves documents and places their text into an LLM's context window — chatbots over internal docs, agentic search, customer-support assistants, code/knowledge assistants.