ai-agents-security

LLM | Agentic | Security | Operations in one github repo with good links and pictures.

MCTS (Model Context Threat Scanner) is a local-first security scanner for MCP servers -- static and live tool discovery, multiple analyzers, auditable risk scores, and JSON, SARIF, and HTML output. For authors and platform teams; CI-ready, no cloud API.

Open Agent Trust Infrastructure (OATI) — an open standard for verifiable AI agent identity, delegated authority, policy enforcement, and signed action receipts.

Open-source context intelligence engine for AI coding agents — repository graph, precise code navigation, MCP, impact analysis, memory, and guarded edits.

Adversarial security review for AI skills, repos, MCP servers, and packages before you install them. A read-only scanner plus a five-persona reasoning pass, ending in a GO / GO WITH MITIGATIONS / NO-GO verdict.

Semantic permissions for AI coding agents. Intenter remembers what commands actually do, not just what they’re called.

Local-first AI project memory and control runtime for evidence, privacy, model routing, human approvals and cross-harness handoffs.

Adaptive engineering control plane for AI coding agents — the right specialist depth for every task, with project context and proof-first execution.

🛡️ Official monorepo for SupraWall security plugins and framework integrations. Deterministic zero-trust guardrails for AI agents, including Prompt Injection protection, PII scrubbing, and policy enforcement across Dify, MCP, LangChain, CrewAI, AutoGen, and Vercel AI.

Panel Review: a guardian agent for AI coding decisions. Four frontier models review an agent's riskiest diffs and commits, with local review gates it cannot silently skip. Install: npx @truverifai/init

AegisMCP Enterprise is a dual-layer security gateway and cost firewall designed to sit between AI models (like Claude and Cursor) and execution engines.

LLM Gateway built with LiteLLM for unified access to multiple AI providers through a single OpenAI-compatible API. Features intelligent model routing, failover handling, load balancing, and centralized LLM management.

AgentJam is an open-source, harness-agnostic runtime and governance system for AI coding agents.

Code Python & R du livre « Devenir Orchestrateur IA — du Débutant à l'Expert » : 19 modules, 14 cas métier, 60 activités corrigées. Ebook : da-ia-data.mychariow.co/orchestrateur-ia

Deterministic policy engine for AI agent x402 payments [Python SDK]

Policy enforcement and permission testing for AI agents. Mock enterprise APIs, simulate synthetic user consent, and block unauthorized data access before deployment.

Deterministic policy engine for AI agent x402 payments [TypeScript SDK]

Deeply analyze and refactor an entire software repository to remove unnecessary complexity, AI-generated code patterns, duplication, dead code, excessive abstraction, and poor structure while preserving intended behavior and producing maintainable code for human engineers.

Universal Expert Skill Suite & Dual-Scope Memory Engine for AI Coding Agents (Cursor, Claude Code, Windsurf, Copilot, Antigravity)