AI Security

Koidex helps teams quickly detect and eliminate risks across the software they rely on, including extensions, packages, apps, and models

TweetFeed collects Indicators of Compromise (IOCs) shared by the infosec community at Twitter. Here you will find malicious URLs, domains, IPs, and SHA256/MD5 hashes

Enforce row-level data permissions server-side before an AI agent runs a query

Open-source security automation platform for teams and AI agents

Classifies agent tool calls by risk, gates dangerous actions, and records tamper-evident audits

Runtime control layer enforcing policy, authorization, kill switches, and compliance before agent actions

API for identity, trust, and risk decisions, helping products verify user trust in critical workflows

Inspect agent traffic and mask sensitive data where supported

AI-agent security challenge platform where players break public agent challenges for weekly rewards

AI agent gateway for prompt-injection defense, secret scanning, auditability, and token compression

Astra helps teams build AI agents with strict privacy controls so sensitive data stays protected

AI-assisted security scanning and penetration testing platform for websites, applications, and compliance workflows

AI-native security testing platform with multi-tool orchestration

A security-first desktop AI agent platform for high-compliance environments

Security layer for encrypted permissions, keys, and access control for AI agents

Continuous security validation platform that pentests every deploy, validates exploitable vulnerabilities, and generates remediation PRs

Read-only, fail-closed security research agent for code and cloud infrastructure

Privacy vault for securely supplying personal data and context to AI agents

On-device Windows safety tool that flags suspected synthetic faces and face swaps during live video calls

Run local AI-agent sandboxes on your own Mac, Linux host, or server

A security layer for AI coding agents

Next-generation Gemini models for agents, reasoning, and cybersecurity

Anonymize sensitive data offline before pasting into AI tools

Delegate and follow up on security work in collaboration tools

Control what connected AI apps may do in your accounts

Explain Mac security settings and AI agent integrations in plain language

Approve, block or log risky AI agent actions

Give every AI agent a real identity, inbox, and phone number

Scan AI-built apps for authentication, database, and secret issues

Let AI use credentials while keeping real passwords hidden

Give agents payment and sign-in abilities without exposing real credentials

Everything for pentest. | Penetration testing knowledge base, depositing security methodologies in an AI Agent executable format

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond

Test your prompts, agents, and RAGs. Red teaming/pentesting/vulnerability scanning for AI. Compare performance of GPT, Claude, Gemini, DeepSeek, and more. Simple declarative configs with command line and CI/CD integration. Used by OpenAI and Anthropic

IronClaw is an Agent OS focused on privacy, security and extensibility

The Python Risk Identification Tool for generative AI (PyRIT) is an open source framework built to empower security professionals and engineers to proactively identify risks in generative AI systems

Fine-grained security controls for AI agent tool and API calls