AI Agent Security: Otonom yapay zeka ajanlarının ve MCP protokolünün güvenlik riskleri. Excessive Agency, tool poisoning ve Meta'nın "İki Kural" (Rule of Two) savunma çerçevesi.
-
Updated
Aug 8, 2026
#
excessive-agency
Here are 5 public repositories matching this topic...
Intentionally vulnerable RAG app (LangChain + Chroma + Ollama) demonstrating 4 OWASP LLM Top 10 vulnerabilities: prompt injection, cross-tenant IDOR, excessive agency, sensitive data disclosure. Red-team lab + portfolio piece.
python red-team data-exfiltration ai-security rag security-rese langchain prompt-injection chromadb llm-security ollama owasp-llm-top-10 excessive-agency
-
Updated
Jul 31, 2026 - Python
Hands-on demo of OWASP LLM06 (Excessive Agency) — tool-calling agent hijacked via indirect prompt injection, with layered defenses.
openai ai-security mitre-atlas prompt-injection llm-security owasp-llm-top-10 agentic-ai tool-calling excessive-agency
-
Updated
May 23, 2026 - Python
IDOR / Broken Object-Level Authorization in AI agent tool-calling — a hands-on vulnerable lab. An LLM tool trusts the object id it is handed and skips the ownership check, so any user reads any tenant's data.
appsec ai-agents ai-security idor security-lab broken-access-control bola llm-security tool-calling owasp-llm confused-deputy excessive-agency
-
Updated
Jul 31, 2026 - HTML
Static security auditor for MCP servers. Finds tools that can do irreversible or external damage and verifies each one actually enforces a human-in-the-loop safeguard — not just declares one. AST-only; never runs your code.
python cli security mcp static-analysis ast owasp ai-safety security-tools devsecops ai-security llm agentic-ai model-context-protocol excessive-agency
-
Updated
Jul 14, 2026 - Python
Improve this page
Add a description, image, and links to the excessive-agency topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the excessive-agency topic, visit your repo's landing page and select "manage topics."