#1LLM security, prompt injection, model extraction, adversarial AI, and AI red teaming tools.
Kitploit recommended
vladimir-tokarev-cyera/llama-cpp-security-patchesSecurity patches for unpatched llama.cpp vulnerabilities (CVE-2026-43626 through CVE-2026-43632) — Cyera Research
ilhomjonr/cve-2026-67598CVE-2026-67598 — Emlog Pro: disabled TLS certificate validation in AI assistant (MITM → API-key theft). CWE-295, CVSS 9.1. Reported by @IlhomjonR.
grassplatypus/cve-2026-45033-classCVE-2026-45033 PoC for Claude Code, not Github Copilot. Worked for Haiku 4.5.
agentrust-io/cmcpcMCP: Confidential MCP Gateway. Hardware-attested policy enforcement for MCP tool calls.
pantheon-security/medusaAI-first security scanner. NEW in v2026.7: Claude Code compromise detection — vet .claude/ hooks, permissions & skills before you clone — plus an…
google/sec-geminiSec-Gemini is a cutting-edge AI model designed to enhance cybersecurity capabilities and empower defenders in the ongoing battle against cyber…
huhusmang/awesome-llms-for-vulnerability-detectionThe community's most comprehensive, continuously-updated index of research on Large Language Models for software vulnerability detection — papers…
cisco-talos/deceiveDECeption with Evaluative Integrated Validation Engine (DECEIVE): Let an LLM do all the hard honeypot work!
garagehq/nightcrawlerLocal AI powered red teamer on a phone
mohamed--abdel-maksoud/pisiguardPISIGuard runs entirely in your browser. It spots names, email addresses, phone numbers, credit card numbers, passwords, API keys, and more, replaces…
george0papasotiriou/cve-2026-1337-ai-coding-assistant-prompt-injection-to-sandbox-escapeA prompt injection in a code‑review bot that executes AI‑generated fixes in a sandbox. The sandbox uses a blacklist to prevent dangerous commands,…
theopaid/cve-2026-66421-openclaw-dashboard-stored-xss-via-lastmessage-session-fieldSecurity Advisory: Stored Cross-Site Scripting Via Agent Messages Leading To Session Token Theft (openclaw-dashboard)
hunt-benito/pgadmin-ai-assistant-sql-injection-cve-2026-17351-lexer-differential-bypassProof-of-concept demonstrating CVE-2026-17351 SQL injection bypass in pgAdmin 4's AI Assistant via sqlparse/PostgreSQL lexer differential, including…
nccgroup/ml-for-rngsDemonstrates using machine learning to predict random number generator sequences, highlighting cryptographic weaknesses through adversarial analysis.
bishopfox/brokenhillA productionized greedy coordinate gradient (GCG) attack tool for large language models (LLMs)
bishopfox/deephackPoC code from DEF CON 25 presentation
bishopfox/local-llm-ctfA small go harness that uses Ollama to orchestrate LLMs in a restricted process flow
